With or without you: Understanding friendship dissolution from childhood through young adulthood

Abstract

Friendship plays a critical role in well-being across the lifespan, making it important to understand the dissolution of these relationships. Friendships end in many ways and for many reasons, which may result in different social and emotional outcomes. In this review, we examine the rapidly burgeoning literature on friendship dissolution from childhood through young adulthood, detailing current understanding of how common friendship dissolution is, as well as why and how it occurs. Research has documented that during childhood and adolescence, friendships are markedly unstable, although little work has examined friendship stability in adulthood. Characteristics of individuals (e.g., behavior, interpersonal cognition), friendships (e.g., homophily, quality), and the contexts in which friendships are embedded (e.g., friends’ broader social networks) have all been linked to friendship stability. We document the diversity of ways in which friendships end, highlighting differences in timing, communication, and intent. Finally, we discuss circumstances under which dissolving a friendship may be helpful versus harmful. Future work should focus on elucidating friendship processes – that is, the positive and negative interactions that occur between friends – that contribute to dissolution, as well as developing interventions that promote the development and maintenance of high-quality friendships. Other critical research agendas are understanding friendship maintenance and dissolution among youth and young adults from marginalized groups and elucidating the ways in which technology may both strengthen and weaken friendships.

Friendship is a critical relationship across the lifespan (e.g., Hartup & Stevens, 1997; Holt-Lunstad et al., 2010). For children and adolescents, and throughout adulthood, having high-quality friendships is associated with greater well-being (e.g., Bagwell & Bukowski, 2018; Blieszner et al., 2019; Hartup & Stevens, 1997). Of course, friendships end, and for most people, losing a friend is not an unfamiliar experience (e.g., Meter & Card, 2016; Rose, 1984). However, the ways in which and reasons why friendships dissolve vary markedly, as do the implications of losing a friendship. In this article, we review the rapidly expanding literature on friendship dissolution, describing what is known about how stable friendships are, as well as why and how friendships end. We also begin identifying markers of when friendship dissolution may be helpful versus harmful, laying a foundation for interventions designed to promote positive friendship functioning.

Friendship in developmental context

Across the lifespan, having friends is associated with positive outcomes, including greater psychological and physical well-being, as well as greater longevity (e.g., Ajrouch et al., 2024; Blieszner et al., 2019; Holt-Lunstad et al., 2010), likely because of the important functions that friendships serve across development (Bagwell & Bukowski, 2018; Hartup & Stevens, 1997). Children’s friendships are grounded in shared activities and companionship (Hartup & Stevens, 1997). During adolescence, youth begin to spend more time with friends than with family members (Larson & Richards, 1991). The core provisions of friendship also evolve, as loyalty, commitment, and intimacy become increasingly important (Clark & Ayers, 1993). During adolescence, social support also becomes an important provision of friendship. For many young adults, friends become the primary providers of social support, particularly for those not in committed romantic relationships (Carbery & Buhrmester, 1998; Macdonald et al., 2023). Close friendships remain an enduring source of emotional intimacy throughout adulthood (e.g., Carstensen, 1992).

Just as the provisions and functions of friendship evolve across development, developmental factors may also shape whether, how, and why friendships end, as well as the emotional implications of these ruptures. Individual characteristics, such as increases in autonomy, as well as contextual shifts such as school transitions and the expansion of social networks are likely to impact experiences with friends. As such, friendship dissolution must be examined through a developmental lens.

How often do friendships end?

Given the importance of friendships, it is crucial to understand how often these relationships dissolve. Unlike family relationships, friendships are voluntary, making them relatively easier to break (Fehr, 1996). Friendships can be dissolved in two ways (see Bowker, 2011): total dissolutions, in which the friendship ends entirely, and friendship downgrades, in which two friends experience a decrease in closeness. Much of what is known about the stability of friendship comes from research with children and adolescents; far less work has examined the friendships of adults.

To study the stability of youth’s friendships, researchers have often used friendship nominations, asking children and adolescents to identify their friends, and often, to rank these friends by closeness (e.g., identify their best friend). Because friendships are dyadic relationships, many researchers define a friendship as occurring when nominations are reciprocated (i.e., two participants identify each other as friends). To examine stability, researchers collect nominations again later. A complete dissolution occurs when youth no longer identify someone as a friend, and a downgrade dissolution occurs when a friend is still nominated but is ranked less close (e.g., a best friend becomes a close friend; Bowker, 2011).

Studies using friendship nominations have revealed that youth’s friendships show marked instability, with a meta-analytic review documenting that about half of friendships were maintained across two time points (Meter & Card, 2016). Note that many studies stop tracking friendships after they end. Some friendships may dissolve temporarily, to be rekindled later (see Hartl et al., 2015). Although studies using friendship nominations have not reliably differentiated between complete and downgrade dissolutions, there is some evidence that the latter is more common. For example, Bowker (2011) found that 66% of a sample of 11-year-olds reported having a friendship dissolve, with 55% describing a downgrade dissolution and 36% experiencing a complete dissolution. In sum, the friendships of children and adolescents are fluid, and most youth will experience friendships dissolving somewhat or completely.

Research on friendship stability has focused on children and younger adolescents, but there is evidence that young adults experience friendship dissolution as well. For example, Rose (1984) found that 57.4% of young adults interviewed reported losing a close friend in the past five years (see also Lapierre & Poulin, 2022). Theoretically, there are reasons to expect that friendships will become more stable with age (see Poulin & Chan, 2010). For example, the transition from childhood through adolescence to early adulthood corresponds to the development of skills that may facilitate effective navigation of challenges in friendship, such as emotion regulation and perspective taking (Wood et al., 2018). As such, it was surprising that Meter and Card’s (2016) meta-analysis found no association between age and stability. This result may have occurred because the studies analyzed predominantly sampled children and early adolescents; thus, participants may have been too young to detect age-related differences. Alternatively, it may be that friendship stability does not actually increase through young adulthood, perhaps because later adolescence and young adulthood are characterized by transitions that may stress friendships. For example, high-school friendships may dissolve as friends graduate and move apart (Oswald & Clark, 2003). Although more research on friendship stability during adulthood is needed, friendship dissolution is clearly a common experience for children, adolescents, and young adults.

Why do friendships end?

In addition to knowing how common friendship dissolution is, it is also important to chart the reasons why friendships end, as the factors precipitating dissolution may have implications for well-being. Studies using friendship nominations have provided insight into why friendships end by examining predictors of friendship instability (e.g., Hartl et al., 2015; see ; Poulin & Chan, 2010). Alternatively, researchers have asked participants to describe or identify the reasons why a friendship dissolved (e.g., Flannery & Smith, 2021; Rose, 1984).

Much of the work examining why friendships end has focused on friendship features, or characteristics of the friendship (Bagwell et al., 2021). Broadly speaking, friendship features can be conceptualized as occurring on one of three levels: (1) friend (individual), (2) friendship (dyadic), and (3) context (see Poulin & Chan, 2010; Figure 1). Individual features are characteristics of each friend that may be associated with the success of the friendship, such as interpersonal skills. Friendship features are properties of the dyad, such as the quality of the friendship. Similarity between friends is also a friendship-level feature. Although gender is an individual characteristic, whether two friends are the same or different genders is a feature of the relationship. Finally, friendships are embedded in broader contexts that may facilitate or strain the relationship.

Figure 1.

Conceptual model liking friendship features, friendship processes, and friendship maintenance and dissolution.

Friendship features

Individual characteristics

Gender

Much work has focused on gender differences in friendship experiences, with studies hinting that associations between gender and friendship stability may vary across developmental periods. Research with children and younger adolescents suggests that friendship stability does not differ as a function of gender. Notably, the meta-analysis by Meter & Card (2016) documented no differences in friendship stability between girls and boys. Similarly, when Flannery and Smith (2021) asked a large sample of early adolescents how many friendship dissolutions they had experienced, boys and girls reported comparable numbers (but see Bowker, 2011). Less work has examined friendships in later adolescence and young adulthood. Rose (1984) found that more women than men reported having a friendship end in both high school and university. Similarly, Lapierre and Poulin (2022) documented greater instability in the friendships of women than of men between 22- and 26-year of age. More research is needed before drawing strong conclusions, but it may be that in young adulthood, women may be more likely to have friendships dissolve than are men.

Participation in romantic relationships

Increasing investment in romantic relationships may also weaken friendships. Recent work found that adolescents in a dating relationship were not more likely to dissolve friendships than were adolescents who were not, perhaps because adolescents’ romantic relationships are often more casual, and so may not detract from their friendships (Zhang & Felmlee, 2024). In contrast, for many young adults, romantic relationships become the primary source of social support, which may reduce investment in friendships (see Langheit & Poulin, 2022). Carstensen (1992) found that during young adulthood, interactions with close friends decreased, possibly in part because more time was being spent with romantic partners. This change did not translate into diminished emotional closeness, but Galambos et al. (2018) found that young adults who were married or living with a romantic partner reported feeling less supported by friends.

Although greater investment in romantic relationships may make friendships more vulnerable, Doherty (2021) highlights the complex ways in which romantic relationships may intersect with friendship during adulthood. Women who took part in this study often described the escalation of their own or a friend’s romantic relationship as a significant turning point in their friendship. Sometimes, friends grew closer – for example, through participation in weddings – whereas other times, they grew more distant.

Attachment

Research has documented associations between attachment style and dissolution of romantic relationships (see Lee & Sbarra, 2013). Although little work has examined attachment styles in the context of friendship, there is some evidence that they may also contribute to dissolution of these relationships. For example, Gillath et al. (2017) explored whether young adults’ attachment style was associated with perception and management of their friendship networks, finding that greater attachment anxiety and avoidance were each associated with stronger endorsement of friendship dissolution. In general, attachment styles may shape communication styles, conflict resolution strategies, and emotion regulation (Feeney & Noller, 1992), which likely has implications for friendships as well as romantic relationships.

Psychological symptoms

Work with youth has linked both internalizing and externalizing symptomatology to friendship stability. Several studies with children and adolescents have documented that greater depressive symptoms are associated with less stable friendships (e.g., Chan & Poulin, 2009; Marengo et al., 2018; Rose et al., 2011), although other work suggests more complex relations between depressive symptoms and friendship functioning (see Schwartz-Mette & Rose, 2016). The links between anxiety symptoms and friendship stability are also not straightforward. Rose et al. (2011) found that greater anxious symptoms were linked to more stable friendships, whereas Marengo et al. (2018) found no associations. Associations between anxious symptoms and stability may depend on the type of anxiety – social anxiety, in particular, may compromise friendship functioning (e.g., La Greca & Lopez, 1988) – as well as symptom severity. Externalizing psychopathology has also been linked to friendship instability, with research showing that the friendships of both children with attention deficit hyperactivity disorder (Spender et al., 2023), and adolescents who are persistently antisocial (see Piehler & Dishion, 2007), are of shorter duration than those of their peers.

Interpersonal behavior

Theoretically, the ways in which individuals behave should be linked to their adjustment in friendship; for example, youth with poor social skills may have difficulty maintaining friendships with many different peers (see Wei et al., 2024). Correspondingly, studies have documented associations between overall behavioral tendencies and friendship stability. As reviewed by Poulin and Chan (2010), greater aggression has been linked to less stable friendships, whereas greater prosocial behavior is associated with more stable friendships. The friendships of children who engage in greater withdrawal – that is, avoiding others by isolating oneself from peers or withdrawing in social interactions – are not less stable than those of their peers (see Rubin et al., 2009). However, recent work found that greater withdrawal at age 16 years was associated with less stable friendships (Barzeva et al., 2022), a developmental difference that may be occurring because withdrawn behaviors are evaluated more negatively by older youth (see Rubin et al., 2009).

Interpersonal cognition

There is also evidence suggesting that interpersonal cognition may be linked to friendship stability. For example, research has shown that individuals have implicit beliefs about the potential of their relationships, which may guide interpretations of relationship experiences and shape responses to relationship challenges (see Knee, 1998). People who hold strong destiny beliefs think that potential relationship partners are naturally compatible, and if they are not, nothing can be done to change that. In contrast, those who hold strong growth beliefs think that relationships can be improved with effort. Destiny and growth beliefs have been linked to behavior with romantic partners (e.g., Knee, 1998), with a meta-analysis showing stronger destiny beliefs were associated with greater likelihood of relationship dissolution (Le et al., 2010). Santucci et al. (2022) extended this work to the friendships of young adults, finding that greater destiny beliefs about friendship were associated with stronger endorsement of ending a friendship in response to hypothetical challenges. In contrast, stronger growth beliefs were associated with greater endorsement of maintenance strategies.

Dyadic

Homophily

A key dyadic predictor of friendship stability is homophily, or similarity between friends (see Laursen, 2017). Yang et al. (2021) found that friendships in which both adolescents were shy were more vulnerable to break up, indicating that sometimes similarity may pose a risk for dissolution. In general, however, greater similarity appears to promote friendship continuity. Greater differences between friends on characteristics such as social competence, victimization, aggression, and peer acceptance have been linked to greater likelihood of the friendship dissolving (e.g., Hartl et al., 2015; Poulin & Chan, 2010; Wei et al., 2024). Similarly, Guimond et al. (2019) found that the more adolescent friends differed in their depressive and anxious symptoms, the more likely the friendship was to end. Notably, individuals’ internalizing symptoms did not predict dissolution, suggesting that it was similarity between friends, not individual functioning, that contributed to friendships’ demise. Although many studies collect data from only one friend (see Bagwell et al., 2021; Dryburgh et al., 2022), this finding highlights that a complete understanding of dissolution requires collecting information about the dyad. The extent to which individuals’ characteristics are associated with friendship dissolution are likely shaped by characteristics of their friends.

Studies have also examined whether same-gender friendships are more stable than mixed-gender friendships, yielding mixed results. Hartl et al. (2015) followed friendships that were newly formed at the start of grade seven through the end of high school, finding that mixed-gender friendships were more likely to dissolve than same-gender friendships. In contrast, Nielson et al. (2020) found that over the course of sixth grade, mixed-gender friendships were as stable as same-gender friendships. The discrepancy in findings may be due to differences in the length of time for which friendships were followed.

Friendship quality

Another important dyadic feature is friendship quality, or the provisions and interactions that characterize the relationship (Asher & Weeks, 2018). For both youth and adults, friendships comprise positive features, including companionship, support, and intimacy, as well as negative features, the most studied of which is conflict, or the extent to which friends disagree and fight. High-quality friendships are high on positive features and low on negative features. Note that although friendship quality is conceptualized as a feature of the relationship, it is often measured as an individual-level characteristic, by assessing one person’s perception of the relationship.

Research with children and adolescents has linked friendship quality and stability. For example, Hiatt et al. (2015) find that among young adolescents, high-quality friendships are quite stable, with 90% staying together across two school years. In contrast, fewer than half of lower quality friendships were maintained over the same period. Recent evidence suggests that lack of positive features in a friendship is more predictive of dissolution than presence of negative features (Faur et al., 2024). To date, the associations between friendship quality and dissolution have not been examined with young adults, but it would be expected that high-quality friendships would be more stable.

Context

The contexts in which friendships are embedded may also be linked to stability. Friendships may form when two people have close physical proximity to each other, such as when two children are in the same class, or two adults work together. Physical separation may hasten the demise of the relationship. When Rose (1984) asked young adults to describe why a friendship had ended, they often described a change that increased distance, such as moving. For children and younger adolescents, changing classrooms each year may result in changes to friend groups (e.g., Hardy et al., 2002).

There may be developmental differences in the extent to which friendships are vulnerable to separation, such that the friendships of children and younger adolescents may be more susceptible to physical distance. Compared to older youth and adults, they will be less able to travel independently and may also have less reliable access to a cellphone or a computer, both of which facilitate virtual contact. In addition, developmental differences in the provisions of friendship may make it more likely that older adolescents and adults will maintain a friendship despite distance (see Hartup & Stevens, 1997). Specifically, for children, shared activities and play are salient provisions of friendship, and these activities may be more difficult to maintain following a physical separation. In contrast, for adolescents and young adults, intimate self-disclosure becomes increasingly important, and these connections may be easier to maintain using technology.

In addition to physical proximity, other contextual features may contribute to friendship stability. Friendships exist in broader network of relationships, features of which may be linked to dissolution (Felmlee & Sinclair, 2018). For example, friendships in which partners have many friends in common may be less likely to dissolve, both because of greater opportunities to spend time together, as well as the possibility that ending the relationship may result in exclusion from the broader group of friends.

Friendship stability may also be impacted by the broader cultural context in which friendships are embedded (see Wei et al., 2024). For example, cultural variability in relational mobility, or the extent to which individuals can voluntarily form and end relationships, will likely influence friendship stability, with friendships potentially becoming more vulnerable as relational mobility increases (see Schug et al., 2010). To date, little work has examined cross-cultural differences in friendship stability. Wei et al. (2024) recently examined the friendships of 10-year-olds in the United States, China, and Indonesia, finding that over one academic year, children in the US had more stable friendships than did those in the other countries. More work should examine cultural variability in the stability of the friendships of children, adolescents, and young adults.

Friendship processes

Research has documented characteristics of individuals, friendships, and contexts that are associated with friendship stability. What is less clear are the mechanisms by which these features contribute to friendship maintenance and dissolution. What happens – or does not happen – between friends that leads them to develop and maintain closeness, or to withdraw from the relationship altogether? Answering these questions requires studying friendship processes, or the interactions between friends (Bagwell et al., 2021; see Figure 1).

Broadly, friendships may end due to the presence of negative interactions or the lack of positive interactions. Work with youth and young adults suggests that a specific type of negative interaction – friendship transgressions – poses a significant risk to friendships. Beginning in childhood, people have expectations about how friends should behave; friendship transgressions occur when friends violate these expectations (see Asher et al., 1996). Transgressions, which can take different forms, such as betrayal or exclusion, are salient challenges in the friendships of children, adolescents, and young adults (e.g., Flannery & Smith, 2021; Kirmayer et al., 2021; MacEvoy & Asher, 2012). When asked why a friendship ended, adolescents frequently describe transgressions that occurred (Flannery & Smith, 2021). When Khullar et al. (2021) asked young adults how they would respond to challenging situations in their friendships, transgressions by a friend were most likely to elicit responses involving dissolving or downgrading the friendship.

Friendships may also end because of a lack of positive interactions (see Flannery & Smith, 2021). Companionship – spending time together and having fun – is an important provision of friendship; not having these experiences may weaken the relationship. When positive interactions are an expected part of a friendship, their absence may constitute a transgression. For example, friends are supposed to share private details of their lives with each other (Asher et al., 1996); failing to do that may constitute a transgression (Kirmayer et al., 2021). In such situations, lack of positive interactions may translate into negative interactions; for example, refusing to help a friend may lead to a fight.

Links between friendship features and processes

Understanding friendship processes may provide key new insights, as they may be the mechanism through which different friendship features influence dissolution and maintenance. At the individual level, an examination of friendship processes may help to clarify conflicting findings regarding gender differences in friendship stability. As reviewed by Rose and Asher (2017), research with children and adolescents has shown that, compared to those of boys, the friendships of girls are characterized by greater caring, affection, intimacy, and helping, differences that may also be present in young adulthood (see Langheit & Poulin, 2022). These differences suggest that the friendships of girls and women should be more stable than those of boys and men; yet studies with children and adolescents have not revealed these expected gender differences (Meter & Card, 2016). Rose and Asher (2017) propose that this paradox may occur because boys and girls excel at different tasks of friendship. Specifically, while girls and women may manage tasks such as self-disclosure and emotional support better than do boys and men, they may struggle more to navigate transgressions in friendship. Girls and women have higher expectations for friendships than do boys and men (Hall, 2011), which may be why girls react more negatively than do boys to hypothetical friendship transgressions (MacEvoy & Asher, 2012), and women tend to be angrier than men when conflicts arise (Benenson et al., 2014). Given that these gender differences in the processes and provisions of friendship are evident in childhood, adolescence, and adulthood, it is unclear why research with children and early adolescents has not documented differences in the stability of girls’ and boys’ friendships, yet research with young adults has suggested that gender differences may exist. Limited work with young adults, as well as methodological differences between studies conducted with children and adolescents and those focused on young adults, make developmental comparisons tenuous at this time.

Careful examination of friendship processes may also elucidate links between psychological symptoms and friendship dissolution, which may be driven, at least in part, by interactions occurring between friends. Greater depressive symptoms have been linked to several behaviors that may generate negative exchanges, or dampen positive ones, between friends. One example is excessive reassurance seeking, which involves repeated bids for reassurance from close others that are subsequently discounted (Starr & Davila, 2008). Prinstein et al. (2005) found greater reassurance seeking by adolescents was prospectively associated with their friends perceiving fewer positive features in their relationship, which may weaken the friendship. Observational work with adolescents has shown that for youth experiencing greater depressive symptoms, greater conversational self-focus – redirecting conversations towards the self and away from others –is linked to greater rejection by friends (Schwartz-Mette & Rose, 2016). The documented associations between these behaviors and interpersonal rejection and friendship quality suggests that they may contribute to friendship dissolution, although few studies have tested this explicitly. Given evidence that greater discrepancies in friends’ depressive symptoms predicts the dissolution of the friendship (Guimond et al., 2019), future work should also consider these behaviors in the context of the depressive symptomatology of both friends. It may be that these behaviors are received more negatively by friends who have fewer depressive symptoms themselves.

More generally, the behaviors in which friends engage likely contribute to positive and negative interactions. Prosocial overtures towards friends – for example, giving compliments – may facilitate positive interactions, whereas aggressive behaviors may escalate conflict. Interpersonal cognition, including beliefs about friendship, may also shape interactions. For example, stronger growth beliefs may lead to constructive conflict resolution strategies, which may ultimately reduce fighting (see Santucci et al., 2022).

At the dyadic-level, homophily – or similarity between friends – may facilitate positive interactions. Similarity in interests may enhance participation in shared activities. Certainly, diverging interests are commonly reported as reasons for dissolution across development (Johnson et al., 2004; Rose, 1984), possibly because there is less opportunity for positive interactions. Similarity may also reduce conflicts (see Hartl et al., 2015). For example, sharing similar political views may contribute to fewer arguments.

Everyday interactions between friends may also connect friendship quality and friendship stability. Friendship quality reflects friends’ understanding of and expectations of the relationship (see Bukowski et al., 2020), and it may be linked to interactions between friends in two ways (see Hinde, 1995). Interactions may shape relationship quality. Friends who fight a lot may come to view their friendship as high in conflict, whereas those who provide consistent emotional aid may evaluate their friendship as supportive. Perceptions of friendship quality may also shape subsequent interactions, and as such, friendship quality may contribute to stability, at least in part, through the behaviors in which friends engage. For example, a friend who views a relationship as caring may be more likely to initiate positive interactions with their friend, and less likely to engage in ways that may escalate conflicts and fights. The perceived quality of the relationship may also influence how friends evaluate and respond to each other’s behavior. For example, Macdonald et al. (2023) recently found that friends who viewed their relationship more positively also reported feeling more supported following an interaction, even after accounting for the specific behaviors that occurred. A complete understanding of the links between friendship quality and dissolution will require mapping the ways in which quality is linked to positive and negative interactions.

Finally, contextual features may also impact interactions between friends. For example, social network composition may shape everyday interactions, for better and for worse. Having friends in common may create opportunities to participate in shared activities. On the other hand, the expansion of social networks that occurs in adolescence and young adulthood (Wrzus et al., 2013) may lead to friends spending time with others, a potential source of conflict (Flannery & Smith, 2021). Overall, understanding how friendship features translate into everyday interactions with friends will provide new insight into the complex dynamics of friendship, opening new windows onto why friendships succeed or fail.

How do friendships end?

Friendships end in many ways. For example, two friends may no longer have any classes together and gradually drift apart. Alternatively, someone may be so angry with their friend that they tell them their friendship is over. How friendships end may have emotional and social implications (see Flannery & Smith, 2021; Vieth et al., 2022), making it important to map the variety of ways in which friendships dissolve.

The examples in the previous paragraph reflect what Vieth et al. (2022) describe as passive and active pathways to dissolution. The former describes a process in which a friendship dissolves gradually. In contrast, active dissolution involves an abrupt, direct termination. This taxonomy highlights two key differences in the ways in which friendships may dissolve. The first is timing: does the friendship end immediately or diminish gradually? The second is whether there is explicit communication about the dissolution of the friendship. Extending this framework, a third factor to consider is intent, or the extent to which someone consciously holds the goal of dissolving the friendship or diminishing its closeness. Someone who is not putting effort into maintaining a friendship may not actively be trying to end the friendship.

These three dimensions – timing, communication, and intent – provide a helpful framework for organizing the ways in which friendships end (see Table 1). Research has documented that youth and young adults describe choosing to terminate their friendships abruptly, both in their real lives and in response to hypothetical scenarios. For example, Flannery and Smith (2021) asked adolescents to identify the ways in which their friendships had dissolved and found that participants reported taking actions to end the friendship immediately. Similarly, Kirmayer et al. (2021) asked young adults how they would manage challenging situations occurring with a friend, noting that participants sometimes said they would immediately sever ties with the friend.

Table 1.

Ways in which Friendships End as a Function of Timing, Communication, and Intent.

Ways in which friendships end	Description	Timing	Explicit communication to friend	Intent to dissolve or downgrade the friendship
Direct rejection	Telling a friend that the friendship is over or that you no longer like them	Immediate	Yes	Yes
Mistreatment	Treating a friend badly to communicate that the friendship is over	Immediate	No	Yes
Ghosting/Avoidance	Avoiding contact with a friend without explaining why	Immediate	No	Yes
Third-party	Having a third-party tell a friend that the friendship is over	Immediate	Yes	Yes
Active distancing	Acting in ways that decrease the frequency of interactions or contact with a friend	Gradual	Yes/No	Yes
Compartmentalizing	Putting limits on a friendship, for example, by restricting the activities one does with a friend or the topics that are discussed	Gradual	Yes/No	Yes
Lack of maintenance	The friendship diminishes in closeness and frequency of interactions decreases	Gradual	No	No

Youth and young adults use different strategies to cut ties with a friend swiftly and intentionally, which vary in terms of whether it is explicitly communicated that the friendship has ended. In a direct rejection, one friend clearly tells the other that the friendship is over (Flannery & Smith, 2021). Alternatively, and less commonly, a friend may indirectly communicate that the friendship is over by treating the friend badly (Apostolou, 2023; Baxter & Philpott, 1982; Flannery & Smith, 2021). Both strategies involve interaction with the friend. In contrast, “ghosting” involves “suddenly ignor[ing] or stop[ping] communicating with another person, without telling them why” (Kay & Courtice, 2022, p. 408). Ghosting behaviors include not responding to text messages, phone calls, or emails, blocking or unfollowing someone on social media, as well as avoiding in-person encounters and ignoring the person. Such avoidant strategies are common. For example, when Flannery and Smith (2021) asked adolescents how their friendships had ended, avoiding the other person was endorsed nearly half the time (see also Baxter & Philpott, 1982). In contrast, direct rejection was reported in approximately 20% of dissolutions. This study also indicated that adolescents may have a third-party relay the message that the friendship is over, a strategy less commonly reported than avoidance and direct rejection.

These strategies all involve a rapid, intentional break. Sometimes people will end their friendships gradually, by acting in ways that reduce the closeness of the relationship without rupturing it completely. When Kirmayer et al. (2021) asked young adults how they would respond to challenging situations with friends, participants sometimes said they would try to distance themselves from the friend (e.g., talk to them less often). Participants also described strategies that involved putting limits on, or compartmentalizing, the friendship, such as restricting the activities they do with the friend (see also Apostolou, 2023). When someone decides to distance or compartmentalize, they may or may not discuss that with the friend. There is evidence that, at least for young adults, these more gradual strategies are more frequently used than those involving an abrupt termination. Khullar et al. (2021) found that young adults reported distancing and compartmentalizing more frequently than they did ending the relationship in response to hypothetical friendship challenges (see also Apostolou, 2023).

The outcome of these more gradual strategies will likely be to diminish the closeness of the relationship, yet a variety of motivations may underlie their enactment. The actor may want to end the friendship and choose to do so by placing restrictions on the relationship (see Vieth et al., 2022); however, distancing may occur without deliberate intent to sever, or even downgrade, the friendship. For example, if one friend moves away, it may be challenging to find ways to connect, which may gradually lead to dissolution over time. After interviewing young adults about the dissolution of their friendships, Rose (1984) noted that “many friendships seemed to deteriorate past the point of being salvaged without [participants] noting their decline” (p. 273). Thus, friendships sometimes end due to passive lack of maintenance, rather than active distancing.

Although active distancing and compartmentalizing strategies involve intent to downgrade the friendship, it is possible that they are enacted as a way of ultimately preserving the relationship. For example, in the aftermath of a betrayal, someone may choose to minimize contact with their friend as a way of managing their negative affect so that the friendship may continue. Although little work has examined this question, available evidence is inconsistent with this hypothesis. Khullar et al. (2021) examined the associations between the goals young adults endorsed in challenging situations with friends and their reported use of dissolution strategies. They found that stronger endorsement of the goal of maintaining the friendship was associated with lower likelihood of reporting distancing strategies and was unrelated to reported use of compartmentalizing.

Lastly, it is important to consider the extent to which the end of a friendship is mutual. In some instances, one friend may choose to end the friendship and in others, the friendship may dissolve through a mutual decision, or both friends may let the relationship fade away. Sprecher et al. (2010) proposed that romantic breakups are typically initiated by one partner. However, dissolution may look different in friendship. Flannery and Smith (2021) found that many of their participants (58.8%) described the end of their friendship as mutual, and suggested that friendships may be more likely to dissolve mutually than romantic relationships because ending a friendship does not require the same explicit breaking up process. Without reports from both friends, it may be difficult to ascertain whether reports of mutual breakup are face saving reports by a friend who has been rejected.

In sum, friendships end in many ways. Future work examining consequences of these different pathways would provide insight into whether some types of dissolution pose particular socioemotional risks, as well as if there are optimal ways for someone to end a friendship. In conducting this work, it will be important to attend to friendship characteristics and processes, which may moderate the links between dissolution pathways and outcomes. For example, a best friendship that dissolves through one friend’s passive lack of maintenance will likely have greater emotional consequences for the partner than if a more casual friendship dissolved in the same way.

Developmental differences in friendship dissolution

As discussed by Vieth et al. (2022), the ways in which friendships end are likely shaped by features of the friendship, at each of the three levels. For example, changes in contextual features, such as one friend moving away, may be more likely to result in gradual, unintentional dissolutions. Although little work has examined developmental differences in the ways that friendships end, it seems likely that strategies will change as a function of developmental level. Baxter and Philpott (1982) asked participants how they would end a friendship, finding that 10-year-olds described fewer types of strategies than did adolescents and young adults. Moreover, 10-year-olds reported proportionally more strategies involving directly communicating that they no longer liked the person. Of course, this work does not cover all possible strategies, however it suggests that the ways in which friendships end may look different with age.

Developmental differences in friendship dissolution may arise for many reasons. Some differences are likely due to individual development. Relational aggression, behaviors that harm by targeting someone’s relationships, are more common during early adolescence than earlier in childhood (e.g., Vaillancourt et al., 2007). Thus, adolescents may be more likely to end a friendship by treating the friend badly, possibly as a way of establishing dominance in their social networks (see Mayeux, 2014). Conceptualizations of friendship may also change across development in ways that shape dissolution. For example, children may view friendship in more absolute terms, such that someone is either a friend or not. As a result, they may be more likely to engage in abrupt termination strategies. As social networks expand and relational experiences accrue, conceptualizations of friendship may become more nuanced, allowing for varying degrees of closeness. This richer understanding may make people less likely to take the step of swiftly ending a friendship. Some developmental differences in friendship dissolution strategies may also result from contextual differences (Hardy et al., 2002). If two friends are in the same class, it may be difficult to end a friendship purely through avoidance; proximity may necessitate a direct conversation. As children move into adolescence, schools get larger and the contexts in which friendships occur expand. For all these reasons, it may be that with greater age, friendships may be more likely to dissolve by fading away, and it may be easier to enact avoidant strategies.

Understanding when friendship dissolution may be helpful versus harmful

The documented benefits of high-quality friendships make it essential to understand friendship dissolution. When friendship dissolution results in someone feeling less supported, it may cause harm (Lessard & Juvonen, 2018). On the other hand, maintaining a friendship may sometimes be detrimental. People remain in friendships characterized by pronounced negativity (Bushman & Holt-Lunstad, 2009), and research has documented prospective associations between having a friendship high on negative features and greater loneliness and depression (Schwartz-Mette et al., 2020). When a particular friendship causes distress, ending it may be beneficial (Flannery & Smith, 2017). Although adolescents commonly report negative emotions, including sadness and anger, when a friendship ends, they often also report feeling relief and happiness (Bowker, 2011; Flannery & Smith, 2021), suggesting that the end of a friendship can be a positive change. To best support youth and adults to maintain high-quality friendships, it is critical to disentangle when friendship dissolution may be helpful versus harmful.

Although researchers have focused on the benefits of friendship, friendships can also cause harm. A growing body of work documents that children, adolescents, and young adults experience physical, verbal, and relational victimization at the hands of friends (e.g., Crick & Nelson, 2002; Daniels et al., 2010; Dryburgh et al., 2023; Dryburgh et al., 2025). In fact, friendship may be a particularly risky context for aggression. Faris et al. (2020) analyzed reports of victimization and aggression in a sample of more than 6000 adolescents, finding that aggression was three times more likely to occur between individuals who were friends than between those who were not. Unsurprisingly, greater victimization in friendship has been linked to poorer mental health outcomes (e.g., Crick & Nelson, 2002; Faris et al., 2020), even after accounting for victimization in other relationships (Dryburgh et al., 2025).

When victimization is occurring, the best choice may be to end the friendship; however, a qualitative study by Bouchard et al. (2021) highlights many reasons that youth may find it difficult to do so, which may also be relevant for adults. People who have few friendships may feel that they must stay in the relationship or risk being friendless. People may also stay in the friendship because of benefits it is conferring. For example, Daniels et al. (2010) found that 9- to 12-year-old girls who reported high levels of relational victimization in a friendship also perceived these relationships to be high on companionship and intimacy. Paradoxically, the presence of positive features may heighten the risk the friendship poses. For example, although intimacy is a positive provision of friendship, greater disclosure also increases vulnerability to betrayal (Faris et al., 2020), as friends may share that information with others.

An additional reason that someone may elect to stay in a friendship involving victimization is that they do not conceptualize the friend’s behavior as atypical (Bouchard et al., 2021), believing that being treated poorly is part of friendship. Conflict is inevitable in close relationships, and in some cases, victimizing behaviors may reflect difficulty managing the intense emotions that can result (see Dryburgh et al., 2023). Sometimes, however, these behaviors may be part of a problematic pattern, such as trying to establish dominance, or they may reflect underlying emotional or behavioral challenges. Disentangling the different reasons for victimizing behaviors within friendship is critical for the development of programs to support people navigating potentially harmful friendships, as well as for prevention and intervention programs targeting the maintenance of high-quality friendships.

Sometimes, people may stay in a friendship too long. Other times, though, people may leave friendships too early, ending a relationship when challenges could be overcome. Given the well-documented benefits of friendship, having a friendship end may confer risk to one’s well-being. These risks may vary as a function of developmental level. Some research suggests that for children and younger adolescents, what may be most important is having a friend, rather than maintaining the same friendship. For example, Chan and Poulin (2009) tracked the stability of the friendships of 12-year-olds across monthly intervals, finding that changes in the identified best friend did not prospectively predict depressive symptoms, after controlling for earlier symptoms. Bowker et al. (2006) examined friendship nominations made by a large sample of 10-year-olds at the beginning and the end of a school year. They documented no differences in socioemotional adjustment between children who maintained the same best friendship over the school year and those who had a different best friendship at the two time points. In contrast, compared to these groups, children who did not have a best friendship at either time point, as well as those who had a best friend at time one but not time two, were more victimized and less prosocial. Thus, changes in the identities of close friends may sometimes reflect typical fluctuations and may not always be cause for concern.

For older adolescents and adults, there may be more marked benefits associated with maintaining the same friendships. With age, it may become more difficult to develop close friendships, which may increase the cost of a friendship dissolving. Carstensen (1992) documented that from late adolescence (age 17 years) to young adulthood (age 30 years), people winnow their social networks, increasingly investing their efforts in a smaller group of family members and close friends, a pattern consistent with data indicating that social networks, and number of friends, decrease after young adulthood (Wrzus et al., 2013). As a result, it may be more difficult to initiate new friendships, as well as potentially challenging to develop new friendships that provide comparable closeness and intimacy to longer-standing relationships.

Future directions

Interventions to support positive friendship adjustment

Given the benefits of close friendships, as well as the potential risks that they pose, a key next step is the development of interventions that promote the maintenance of high-quality friendships. Intervention focus will depend on both developmental level and the specific friendship challenges being experienced. Youth and adults who have difficulty initiating new friendships may need support to develop those skills. Some youth and adults may also benefit from guidance regarding whether the harms being caused by a friendship outweigh the benefits, as well as from support during the dissolution process. For adults, in particular, there may be a need to develop interventions that promote the maintenance of specific, on-going friendships. As social networks constrict, it may become more difficult to initiate new friendships. Moreover, theoretically, people who have been friends longer will enjoy a richer co-constructed history that may contribute to greater closeness, making these friendships harder to replace. For these reasons, friendship interventions with adults, as well as with older adolescents, may benefit from a focus on developing their existing friendships, including effective navigation of conflicts and transgressions that have occurred. Given the fluidity of the friendships of children and younger adolescents, focusing on the maintenance of specific friendships may be less critical, but it will be important for youth to develop and practice the skills necessary to maintain high-quality friendships.

Some youth and young adults may also benefit from brief interventions targeting implicit beliefs about friendship. There is some evidence that people endorsing stronger destiny beliefs may be more likely to end a friendship when challenges occur, whereas those endorsing stronger growth beliefs may be more likely to engage in maintenance behaviors (Santucci et al., 2022). Targeting those beliefs may result in changes to dissolution and maintenance behaviors. Although little work has examined the malleability of implicit theories of relationships, other types of implicit theories (e.g., personality) can be altered with single-session interventions (see Schleider & Weisz, 2018).

Strengthening growth and weakening destiny beliefs about friendship may be useful for adolescents and young adults who dissolve relationships frequently. Although some turnover in friendships is expected, some research suggests that greater instability in friendship networks – characterized by consistently dissolving friendships and forming new ones – is associated with poorer adjustment, including greater depressive symptoms (Chan & Poulin, 2009) and poorer academic functioning (Lessard & Juvonen, 2018). Targeting implicit beliefs may contribute to greater stability by helping youth and young adults to view challenges in the friendship as opportunities to develop the relationship, rather than as indicators that this friendship was not meant to be.

Understanding the role of technology in friendship maintenance and dissolution

There is also a need for more research focused on the role of communication technology and social media in the maintenance and dissolution of friendships. Technological developments – including the widespread adoption of cellphones, greater accessibility of texting and instant messaging, and increased availability of video calls – have profoundly changed the ways in which friends interact. Many of these developments are positive. For example, use of these technologies may facilitate friendship maintenance when contextual changes occur; for example, when a friend moves away (see Cummings et al., 2006). Similarly, social media has introduced new opportunities to connect with friends, interactions that may serve to strengthen relationships (e.g., Décieux et al., 2019). Technological advancement has also introduced new challenges, such as the potential for jealousy when one sees friends interacting with others online (Vaillancourt et al., 2024). For adolescents and young adults, much of their communication with friends is technologically mediated (see Anderson & Jiang, 2018; Lenhart et al., 2015; Scott et al., 2021), making it important to understand what skilled communication looks like in these modalities, as well as the ways in which technology both facilitates and threatens friendships.

Friendships of individuals from marginalized groups

Finally, it is critical that more research focus on the friendship experiences of those from marginalized groups, including youth and young adults with sexual and gender minority identities, as well as those with BIPOC identities. Most work on the friendships of children, adolescents, and young adults has been conducted with White, North American samples (see Dryburgh et al., 2022; Fehr & Harasymchuk, 2022; Meter & Card, 2016). Moreover, research has often examined same-gender friendships, both because most of youth’s and young adults’ friendships are with same-gender peers (see Mehta & Strough, 2009) and because of documented differences between the friendships of girls/women and boys/men (Hall, 2011; Rose & Asher, 2017). This focus may exclude gender-minority individuals. As reviewed by Fehr and Harasymchuk (2022), personal identities play a crucial role in friendship maintenance and dissolution, shaping the reasons why friendships end or stay together. The affordances of close friendships may be especially important for individuals from marginalized groups, making it vital to understand factors contributing to the maintenance of high-quality friendships in more diverse samples.

Conclusion

Most children, adolescents, and young adults will have friendships end; however, the emotional and social impact of these experiences will vary. To understand when friendship dissolution poses a risk to well-being, as well as to know how best to support youth and young adults struggling in their friendships, it is crucial to understand why and how friendships end. In this review, we discussed what is known about features of individuals, friendships, and contexts that contribute to the dissolution of friendships, highlighting the necessity of investigating specific processes that may precipitate friendships’ demise. We mapped the ways in which the friendships of youth and young adults may end and began to identify markers that may help disentangle when ending a friendship may be helpful or harmful. Although work on friendship dissolution has burgeoned in recent years, studies have often yielded conflicting findings, and more work is needed to map reliably factors and processes that may strengthen or weaken friendships. Given the benefits that friendship confers, as well as the risks that it can pose, developing tools to promote the development and maintenance of high-quality friendships should be a priority.

ORCID iD

Katya Santucci https://orcid.org/0000-0002-3051-9992