A Quantitative Analysis of Street‐Level Drugs Seized in New South Wales, Australia: Purity, Dose and Adulterants

ABSTRACT

Introduction

The Combined Surveillance and Monitoring of Seized Samples (CoSMoSS) program is a joint harm reduction initiative between NSW Health and NSW Police Force. Here, we examine the contents and purity of a sample of seized suspected target drugs (cocaine, heroin, methamphetamine and 3,4‐methylenedioxymethamphetamine ‘MDMA’).

Methods

Between January 2021 and March 2025, 3378 street drug samples seized during routine operations by NSW Police Force within music festivals or specified geographic regions and suspected by police to contain a target drug were selected for expedited quantitative analysis with GC–MS, with results available within 1 month.

Results

As expected, most samples contained at least one target drug. About 1 in 6 samples (16.5%) were found to contain at least one adulterant and 3.9% of samples contained a novel psychoactive substance, including dipentylone, nitazenes, fentanyl analogues and novel benzodiazepines. Purity information is presented; median MDMA tablet dose was 104 mg and capsule dose was 64 mg (free base) and was higher in more recent years.

Discussion and Conclusions

Some rarer substances with higher or unknown risk potential, including novel psychoactive substance, were detected. Adulteration of cocaine and methamphetamine with opioids and high variability in MDMA purity and dose, remain public health issues. CoSMoSS helps inform harm reduction messaging. Analysis of seized substances provides high‐quality objective information about illicit drug contents and purity, enabling dissemination to the community when high risk substances are identified. These findings can also inform harm reduction messaging and programs and are monitored within the NSW Health drug early warning system.

Key Point Summary

• We present analysis of 3378 street‐level drug seizures over 4 years in New South Wales, Australia.

• There have been more high‐dose MDMA tablet seizures in recent years.

• Adulterants were present in 1 in 6 samples and varied by main drug identified.

• Novel psychoactive substances were present in 3.9% of samples.

• Analysis of street‐level drug seizures forms a valuable part of a drug early warning system for supporting drug warnings and alerting clinicians and informs public health messaging for people who use or may be considering using drugs.

1. Introduction

An effective drug early warning system requires up to date and reliable knowledge of local drug markets, with information from multiple and varied data sources, so that harm reduction advice can be tailored to current risks. One data source which can contribute to an early warning system is chemical analysis of drugs seized by police, such that trends in the contents, purity and adulteration of drugs can be analysed. Few such systems exist in Australia, and the results of analyses are not timely, not typically shared in detail across agencies, nor shared publicly. This paper presents the results of 4 years of analyses of drug samples seized by New South Wales (NSW) Police Force and submitted for analysis as part of the NSW Combined Surveillance and Monitoring of Seized Samples (CoSMoSS) program.

Partly in response to an inquest into six MDMA‐related deaths at music festivals between late 2017 and early 2019, the NSW Government established both the Music Festivals Act and the CoSMoSS pilot project. The Music Festivals Act required organisers of music festivals deemed to carry ‘high risk’ for drug‐related harm to have an agreed health and medical plan to mitigate and respond to risk. Separate to this, the CoSMoSS pilot project focused on quantitative data of street‐level police seizures suspected to be MDMA at high‐risk music festivals over the 2019/2020 festival season. Selected samples underwent analysis of the contents (including adulterants and novel psychoactive substances), purity and dose. Results from the pilot project have been published previously [1]. Later, in response to the Special Commission of Inquiry into the drug ‘Ice’, the CoSMoSS program was expanded to include samples of street‐level substances suspected to contain MDMA, methamphetamine, cocaine or heroin and to include seizures from the community as well as at high‐risk music festivals.

The CoSMoSS program has several strengths: principally, the analysis is timely, providing at least monthly updates to ongoing knowledge about the NSW drug market. Timely information enhances prompt public health responses. Second, the quantitative analysis is objective and can be used in conjunction with survey and other subjective measures [2, 3] to triangulate information on substance identification and purity. Third, the analysis of small‐quantity seizures (which would otherwise not necessarily have been subject to extensive analysis) provides a complementary focus to the large‐scale police seizure data available at the federal level [4] and reflects the contents and purity of drugs at the level of community consumption. Thus, the CoSMoSS program fills an evidence gap informing harm reduction. In its focus on street‐level drug samples, the CoSMoSS program has parallels with front‐of‐house drug checking approaches, which are still in their infancy in Australia, but differs in that samples submitted for drug checking are voluntary, results are provided back to the submitter and harm reduction advice based on the results can be provided. Reports on findings of drug checking approaches in Australia [5, 6, 7, 8] are also not publicly available at the level of detail presented here (e.g., are often presented as snapshots rather than time series).

This paper provides an overview of more than 4 years of quantitative analysis of street‐level drug samples, providing a detailed view of the contents, purity and dose of these drugs. By making this information accessible, we aim to demonstrate that such programs can provide a valuable source of information for a drug early warning system, ensuring the timely delivery of targeted consumer communications for harm reduction.

2. Methods

2.1. Selection of Samples

All of the samples included in this analysis were seized as part of routine operations by NSW Police Force officers starting January 2021, and with analysis completed by March 2025. Samples were selected for inclusion in the CoSMoSS program; Figure 1 describes the sample selection process and constraints. A drug detection refers to what appeared on visual inspection to be a unique instance of illicit tablets, capsules, powder, crystal or liquid found on a single person or in a single instance when drugs were found on the ground. For example, multiple identical tablets would be considered a single seizure, and different samples (different tablets or different forms) seized from the same person would be considered multiple seizures.

FIGURE 1.

Process and constraints of Combined Surveillance and Monitoring of Seized Samples (CoSMoSS) sample selection among all NSW Police Force drug seizures.

The drug types of interest for this project were initially MDMA, methamphetamine, cocaine or heroin, based on their contribution to harms in NSW; later, ketamine was added as a drug of interest where sampling allowed. Selection of samples for further analysis was based on the assessed constitution of the samples made by police at the time of seizure. Police formed suspicions of drug type using their own experience and judgement as well as the available evidence, including any admissions made by the person from whom the drug was seized, the physical appearance of the substance, packaging and the results of presumptive testing where available. Note that the expectation of the person from whom the drug was seized is not recorded.

When selecting the samples to undergo analysis, police limited the samples to those where the suspected drug was one of the drugs of interest (noting that heroin is not often observed among music festival seizures). Police also prioritised detections above a minimum weight (initially, 0.5 g, and later 0.2 g) to increase the likelihood that comprehensive analysis could be conducted, and below a maximum weight (which differs by drug) to avoid including supply‐level seizures. Sample selection is performed separately depending on the source of the seizure.

2.2. Music Festival Samples

As part of the Music Festivals Act, all music festivals meeting certain criteria are assessed for risk; where the festival is deemed to have a high likelihood of drug‐related harm, an agreed health and medical plan for risk mitigation and response is required to be approved prior to the festival. While the language of the guidelines for defining ‘high risk’ [9] changed over the period of interest, the criteria in Section 1.3 of the guidelines remained broadly unchanged, and focused on the target age demographic of the festival, the number of festival patrons, and the style of music (especially high‐energy or electronic dance music). Further factors increasing risk included the event duration, anticipated weather conditions (especially high temperature), and the distance by road to the nearest tertiary health facility. All festivals where CoSMoSS operations took place are selected from these ‘high‐risk’ festivals.

Between January 2021 and March 2025, 66 music festivals in NSW (49 metropolitan and 17 regional) were selected for inclusion in the CoSMoSS program. For context, COVID‐19 related restrictions in NSW continued from March 2020 to October 2021; there were only three CoSMoSS music festivals in 2021. For all years, most (but not all) festivals occur in warmer months (Q1 and Q4). Among all detections as part of routine operations across the 66 music festivals, NSW Police Force exhibits officers sent a sample of 1370 detections (average of 21 per festival) to be analysed as part of the CoSMoSS program. In a few instances, additional samples were selected from a music festival where there was evidence of increased drug toxicity presentations associated with the festival.

It is worth noting that the CoSMoSS music festival selections were independent of the NSW drug checking trial, which operated at selected music festivals in NSW across 2025 and 2026. The first such festival was held in March 2025 and was not a music festival selected for the CoSMoSS program. Therefore, there is no overlap between the data presented here and data associated with the drug checking trial.

2.3. Community Samples

Community samples were seized as part of routine operations by NSW Police Force within a specified geographic area. Introduction of community samples was implemented in a staged approach across geographic areas. Between April and November 2021, samples were selected from two of the six police regions; a third region was added from December 2021 onwards. Metropolitan and regional NSW are represented in the three regions sampled.

Sample selections were made monthly by Crime Prevention Command, utilising established criteria. The incident must be linked to an existing exhibit management record, meaning that there was a substance available for sample, the category must include the word ‘drug’ and the sample needed to be in the weight range described above. The pool of samples eligible for the project did not include large quantity seizures, or samples where the description of the sample did not match the suspected drug (e.g., MDMA described as plant matter).

Once the available sample set to select from was identified, a selection of samples from across the target drug groups was made for further analysis. The number of monthly samples selected for each drug type was not proportionate to the number available to select from. Rather, the aim was to ensure that a minimum count per month was met for each drug type and allow for meaningful analysis of trends over time for each drug type. The minimum samples for each of the target drugs were eight cocaine, seven heroin, six methamphetamine and four MDMA (based on suspected drug type) for each of the participating Police Regions. When there were insufficient samples to select from for a drug type in a given month, another drug type could be over‐sampled up to a maximum sample count of 25 samples per region per month. From June 2023, ketamine was nominated as a drug of choice where additional samples allowed.

2.4. Analysis of Drug Composition

Analyses were conducted by the Illicit Drugs Analysis Unit, NSW Health Pathology Forensic & Analytical Science Service (FASS), where weight, identification, adulterants, purity and dose (where applicable) were determined for each sample. Identification analysis (of drug type) was carried out in line with laboratory procedures and the Scientific Working Group for the Analysis of Seized Drugs guidelines. Gas chromatography–mass spectrometry (GC–MS) was the primary analysis method used for identification, with some combination (depending on the presentation of the sample) of presumptive colour tests, Fourier transform infrared spectroscopy and/or Raman spectroscopy also used for the identification.

In preparation for identification by GC–MS screenings, substances were extracted into an appropriate solvent, with a targeted final concentration of approximately 1 mg/mL. The GC–MS screening did not target specific compounds. If a peak was detected, the mass spectral data were compared to reference libraries containing several thousand compounds, including an extensive range of drugs, poisons and illicit drug cutting agents. Primary (main) illicit drug, along with additional drugs and other adulterants, was reported.

Where practicable, purity analysis was undertaken on each sample confirmed to contain methamphetamine, cocaine, heroin, MDMA or ketamine, to the nearest 0.5%. Between 20 and 120 mg of each homogenised sample was analysed in duplicate using Ultra Performance Liquid Chromatography with a photodiode array detector. Purities are reported as the free base of the drug detected. FASS do not determine salt or base form as part of their workflow. Targeted quantification of lidocaine was introduced for CoSMoSS because of public health concerns from other NSW surveillance data sources [10].

When compared to powders and crystalline substances, tablets comprise additional binding agents (which add impurities) that non‐tablet forms generally do not. Purity within tablets is therefore presented separately from purity within capsules in this paper, and it is noted that dosage is the preferred comparison metric for tablet and capsule forms. Dosages were calculated as a function of the weight and purity (as free base) of seized samples of MDMA. Monthly provision and preparation of Microsoft Power BI reports ensured continuous visibility of outcomes to the stakeholders and that any detections posing a risk to public health could be assessed and managed in a timely manner.

An adulterant was defined as any additional substance detected by FASS. Note that FASS does not test for an exhaustive list of adulterants such as sugars (e.g., lactose) or creatine as many adulterants have very different chemical structures which may not be detectable on drug screening methods. The NSW Health Standing Panel on Toxicity Risk (a panel of clinical and public health experts and community organisation representatives) reviewed identified adulterants and provided advice on their harm potential, based on the adulterant types and their concentration. The same panel also reviewed findings of high purity and/or high dose samples to conduct case specific risk assessment, where required.

2.5. Statistical Analysis

All descriptive analyses were conducted in RStudio, version 2024.04.2. Number and percentage of seized samples were tabulated for music festival and community seizures separately and in total, by main drug detected. Target drugs included cocaine, methamphetamine, MDMA and heroin. Median and interquartile range statistics are reported as measures of central tendency and dispersion respectively due to skewed distributions of purity and dose for target drugs only; boxplots present purity over time (by quarter) for all target drugs (with MDMA separated into tablet, capsule and crystal or powder forms) and dosage over time for MDMA tablets and capsules. Note that the risk of an adverse event increases with MDMA dosage [11, 12]; there is no consensus on what constitutes a ‘high’ dose of MDMA but generally a threshold of 140–150 mg would be considered high [13]. Analyses over time use the date of seizure; the most recent quarter is incomplete, as not all seizures that occurred in the latest months could be analysed by March 2025. Adulterants (drugs present in the sample, but not listed as the main drug) were tabulated by the main drug and detections of novel psychoactive substances were also tabulated.

3. Results

3.1. Main Drugs Detected

Of the 3378 samples analysed, MDMA was the main drug detected in the largest percentage across music festivals and community samples (31.6%) (see Table 1). MDMA was prioritised for sampling in the music festival cohort, so the high proportional representation (61.5%) does not necessarily reflect relative distribution in this setting. Among community samples, the target drugs together accounted for 88.8% of samples, with most detections for methamphetamine. Note that MDMA accounts for a smaller proportion of community than music festival seizures by sampling design. Similarly, heroin was limited to community samples, as it is rarely seized at music festivals.

TABLE 1.

Number and percentage of seized samples from music festival and community seizures, by main drug detected, New South Wales, April 2021 to March 2025.

Main drug detected	Music festival, n (%)	Community, n (%)	Total, n (%)
Target drugs
MDMA	842 (61.5)	225 (11.2)	1067 (31.6)
Cocaine	170 (12.4)	628 (31.3)	798 (23.6)
Methamphetamine	18 (1.3)	705 (35.1)	723 (21.4)
Heroin	0 (0.0)	225 (11.2)	225 (6.7)
None identified a
None identified	71 (5.2)	112 (5.6)	183 (5.4)
Not examined	2 (0.1)	20 (1.0)	22 (0.7)
Missing	1 (0.1)	0 (0.0)	1 (< 0.1)
Other
Ketamine	119 (8.7)	33 (1.6)	152 (4.5)
N,N‐dimethylpentylone	23 (1.7)	15 (0.7)	38 (1.1)
LSD	29 (2.1)	0 (0.0)	29 (0.9)
Dexamphetamine	15 (1.1)	2 (0.1)	17 (0.5)
Cannabis plant matter	10 (0.7)	6 (0.3)	16 (0.5)
MDA	10 (0.7)	3 (0.1)	13 (0.4)
Psilocin	9 (10.7)	0 (0.0)	9 (0.3)
1,4‐butanediol	4 (0.3)	3 (0.1)	7 (0.2)
Isobutyl nitrite	7 (0.5)	0 (0.0)	7 (0.2)
Diazepam	4 (0.3)	2 (0.1)	6 (0.2)
Sildenafil	2 (0.1)	3 (0.1)	5 (0.1)
2C‐B	4 (0.3)	0 (0.0)	4 (0.1)
Bromazolam	3 (0.2)	1 (< 0.1)	4 (0.1)
Lisdexamphetamine	4 (0.3)	0 (0.0)	4 (0.1)
Methandienone	0 (0.0)	4 (0.2)	4 (0.1)
2‐fluoro‐2‐oxo PCE	1 (0.1)	2 (0.1)	3 (0.1)
Amphetamine	2 (0.1)	1 (< 0.1)	3 (0.1)
Chloromethcathinone	3 (0.2)	0 (0.0)	3 (0.1)
Pregabalin	0 (0.0)	3 (0.1)	3 (0.1)
Alprazolam	2 (0.1)	0 (0.0)	2 (0.1)
DMT	0 (0.0)	2 (0.1)	2 (0.1)
Fentanyl	0 (0.0)	2 (0.1)	2 (0.1)
Methylphenidate	2 (0.1)	0 (0.0)	2 (0.1)
O‐acetylpsilocin	2 (0.1)	0 (0.0)	2 (0.1)
Psilocybin	2 (0.1)	0 (0.0)	2 (0.1)
2‐methylmethcathinone	1 (0.1)	0 (0.0)	1 (< 0.1)
Buprenorphine	0 (0.0)	1 (< 0.1)	1 (< 0.1)
Clonazepam	1 (0.1)	0 (0.0)	1 (< 0.1)
Clonazolam	1 (0.1)	0 (0.0)	1 (< 0.1)
Ephedrine/pseudoephedrine	0 (0.0)	1 (< 0.1)	1 (< 0.1)
Fluorexetamine	0 (0.0)	1 (< 0.1)	1 (< 0.1)
GBL	1 (0.1)	0 (0.0)	1 (< 0.1)
GHB	1 (0.1)	0 (0.0)	1 (< 0.1)
Helional amide	1 (0.1)	0 (0.0)	1 (< 0.1)
Morphine	0 (0.0)	1 (< 0.1)	1 (< 0.1)
Olanzapine	1 (0.1)	0 (0.0)	1 (< 0.1)
Oxandrolone	1 (0.1)	0 (0.0)	1 (< 0.1)
Oxazepam	0 (0.0)	1 (< 0.1)	1 (< 0.1)
Oxycodone	0 (0.0)	1 (< 0.1)	1 (< 0.1)
Paracetamol	0 (0.0)	1 (< 0.1)	1 (< 0.1)
Perindopril	0 (0.0)	1 (< 0.1)	1 (< 0.1)
Temazepam	0 (0.0)	1 (< 0.1)	1 (< 0.1)
Testosterone	0 (0.0)	1 (< 0.1)	1 (< 0.1)
THC	1 (0.1)	0 (0.0)	1 (< 0.1)
Tramadol	0 (0.0)	1 (< 0.1)	1 (< 0.1)
Total	1370 (100.0)	2008 (100.0)	3378 (100.0)

Note: ‘/’ denotes that FASS was unable to distinguish between drugs for at least some samples in the period of interest, due to lack of a reference standard.

Abbreviation: THC, tetrahydrocannabinol.

^a For samples where no main drug was identified, additional substances were detected for some samples but are not tabulated here. Samples that are ‘Not examined’ are ones where no tests were performed (e.g., the material was an occupational health and safety risk or was likely to damage the analytical machines).

Approximately 1 in 10 samples (n = 359, 10.6%) had drugs other than the target drugs as their main drug detected, most of which were ketamine. Ketamine's prominence is likely reflective of its addition as a drug of choice where sampling allowed.

3.2. Purity and Dose

Among the target drugs selected from all settings, analysed samples where the main drug detected was methamphetamine (crystal or powder) had the highest median purity (79%), while seized samples of MDMA tablets had the lowest median purity (20%) (Table 2). Note that the range of purity across samples was wide for most drugs.

TABLE 2.

Median (free base) purity and dose from music festival and community seizures, by main drug detected, New South Wales, April 2021 to March 2025.

Main drug in sample	Number of samples	Purity, % (median and range)	Dose a (mg/unit) (median and range)
Target drugs
Cocaine	728	43 (1–91)	—
Heroin	184	72 (8–81)	—
Methamphetamine (crystal or powder)	615	79 (1–86)	—
MDMA (tablet)	171	20 (6–46)	104 (14–216)
MDMA (capsule)	533	74 (5–88)	64 (5–243)
MDMA (crystal or powder)	222	74 (4–82)	—
Other drugs
1,4‐Butanediol	1	93	—
2C‐B	1	2	—
Amphetamine	2	36 (2–71)
Methamphetamine (other)	2	32 (6–57)	—
Pseudoephedrine	1	64	—
MDA	13	8 (4–14)	—
Ketamine	133	70 (1–88)	—

^a Average dose is calculated as: purity/100 x Average weight (mg), where the Average weight (mg) is the Total Weight (mg)/the number of items received. Average dose could not be calculated from MDMA tablet seizures that included damaged tablets, tablet fragments or powder (n = 53). MDMA purity and dose are reported as free base.

Figure 2 shows purity for each drug and form over time. Purity is relatively high and stable for most seized samples where methamphetamine (crystal or powder) is detected. For cocaine, heroin, and all forms of MDMA, there is more variability in purity both within and between quarters over time. Median purity of heroin was high in early 2022, followed by a decline in late 2022 that was maintained until late 2023. Purity has been higher for MDMA capsules since mid‐2023, while MDMA tablets are consistently lower in purity.

FIGURE 2.

Count and distribution of (free base) purity (%) by year and quarter (by date of seizure), and by main drug detected in sample (and form, for MDMA), from January 2021 to March 2025. Music festival and community seizures are included; year and quarter is defined by the date of the police seizure event. The panel groupings indicate the main drug identified in the sample, and statistics were calculated separately for different forms of MDMA. Heavy horizontal bars represent the median, boxes correspond to the first and third quartiles (25th and 75th percentiles), while the whiskers and points show values outside this range. The numbers displayed above each box represent the total number of samples where the target drug was identified as the main drug in a sample and was identified as having purity/dose information available. Purity is presented as free base.

Median dose is calculated only for samples where the main drug was MDMA and was found in either tablet or capsule form and is presented as MDMA free base. From April 2021 to March 2025, the median average dose for MDMA capsules is 64 mg, and was significantly higher for festival samples (65 mg) than community samples (58 mg; Wilcoxon rank sum p = 0.013). For tablets, the median dose is 104 mg, and was not significantly different (p = 0.742) between festival samples (102 mg) and community samples (119 mg). Of 522 capsules for which dose information was available, 6 (1.1%) had an MDMA dosage of 140 mg or more; for tablets, 26 of 121 (21.5%) had an MDMA dosage of 140 mg or more. Median MDMA dose is highly variable within and between quarters for both capsule and tablet forms (see Figure 3); despite this, median dose appears to have slightly increased since the start of 2023 for capsule form. The Supporting Information shows the average weight of MDMA tablets and capsule contents per unit over time, as well as statistical analysis of median purity, dose and weight by year and drug form.

FIGURE 3.

Count and distribution of average MDMA dose (free base), by year and quarter (by date of seizure) and form of MDMA from January 2021 to March 2025. Music festival and community seizures are included; year and quarter is defined by the date of the police seizure event. Heavy horizontal bars represent the median, boxes correspond to the first and third quartiles (25th and 75th percentiles), while the whiskers and points show values outside this range. The numbers displayed above each box represent the total number of samples where MDMA was identified as the main drug in a sample and identified as having purity/dose information available. Dose is presented as free base.

3.3. Adulterants

Of the 3378 samples analysed, 930 (27.5%) were found to contain at least one adulterant (Table 3). Dimethyl sulfone (also known as methylsulfonylmethane) was commonly detected across drugs. It is a common cutting agent, can be used as a health supplement and carries low toxicity potential. A minority of samples (n = 124, 17.2%) where the main drug was methamphetamine were found to contain adulterants; a higher proportion of samples where the main drug was heroin were found to contain adulterants (n = 67, 29.8%). By contrast, cocaine was commonly found to contain adulterants (n = 367, 46.0%). However, the common adulterants in cocaine were lower risk substances including caffeine; lidocaine was commonly detected but the median purity was low (14%). Methamphetamine was identified as an adulterant in samples where the main drug was MDMA and heroin, and heroin was infrequently identified as an adulterant in samples where the main drug was methamphetamine and cocaine (two samples each).

TABLE 3.

Number and percentage of samples containing adulterants, by main drug detected in sample, music festival and community samples, New South Wales, April 2021 to March 2025.

Adulterant	Number of samples containing adulterant (%)
Main drug detected: none identified (n = 183)
Dimethyl sulfone	36 (19.7)
Caffeine	14 (7.7)
Paracetamol	6 (3.3)
Procaine	6 (3.3)
Tiletamine	6 (3.3)
Fluorexetamine/2‐fluoro‐2‐oxo PCE	5 (2.7)
3‐hydroxy‐PCP (3‐hydroxy‐phencyclidine)	2 (1.1)
Diazepam	2 (1.1)
Etizolam	2 (1.1)
Guaiphenesin	2 (1.1)
Lidocaine	2 (1.1)
N‐isopropylbenzylamine	2 (1.1)
N‐cyclohexyl methylone	2 (1.1)
Nitrazepam	2 (1.1)
Xylazine	2 (1.1)
2‐fluoro‐2‐oxo PCE	1 (0.5)
Clozapine	1 (0.5)
Dapoxetine	1 (0.5)
Fluconazole	1 (0.5)
Gabapentin	1 (0.5)
Helional aldoxime	1 (0.5)
Hydroxetamine	1 (0.5)
Levamisole	1 (0.5)
Nimetazepam	1 (0.5)
Orphenadrine	1 (0.5)
Phenacetin	1 (0.5)
Pregabalin	1 (0.5)
Protonitazepyne/isotonitazepyne	1 (0.5)
Tetracaine	1 (0.5)
Main drug detected: missing (n = 1)
Chloromethcathinone	1 (100.0)
Main drug detected: methamphetamine (n = 723)
Dimethyl sulfone	108 (14.9)
N‐isopropylbenzylamine	10 (1.4)
Dimethylamphetamine	4 (0.6)
Amphetamine	2 (0.3)
Caffeine	2 (0.3)
Heroin	2 (0.3)
Chlorphentermine	1 (0.1)
Cocaine	1 (0.1)
Ketamine	1 (0.1)
MDMA	1 (0.1)
Oxazepam	1 (0.1)
Paracetamol	1 (0.1)
Procaine	1 (0.1)
Tiletamine	1 (0.1)
Main drug detected: MDMA (n = 1067)
Dimethyl sulfone	208 (19.5)
Methamphetamine	33 (3.1)
Caffeine	22 (2.1)
N,N‐dimethylpentylone	12 (1.1)
Fluorexetamine/2‐fluoro‐2‐oxo PCE	11 (1.0)
Ketamine	11 (1.0)
MDA	8 (0.7)
MDEA	8 (0.7)
N‐isopropylbenzylamine	7 (0.7)
Chloromethcathinone	2 (0.2)
Cocaine	2 (0.2)
Doxylamine	1 (0.1)
Eutylone	1 (0.1)
Guaiphenesin	1 (0.1)
Ibuprofen	1 (0.1)
Lidocaine	1 (0.1)
MDP2P	1 (0.1)
MDPBP	1 (0.1)
N‐Moc‐MDMA	1 (0.1)
Paracetamol	1 (0.1)
Main drug detected: ketamine (n = 152)
Dimethyl sulfone	21 (13.8)
Tiletamine	8 (5.3)
Caffeine	2 (1.3)
MDMA	2 (1.3)
Amphetamine	1 (0.7)
Cocaine	1 (0.7)
Lidocaine	1 (0.7)
N‐benzyl‐N‐cyclohexylamine	1 (0.7)
Procaine	1 (0.7)
Main drug detected: heroin (n = 225)
Dimethyl sulfone	60 (26.7)
Caffeine	6 (2.7)
Methamphetamine	5 (2.2)
Main drug detected: cocaine (n = 798)
Lidocaine	153 (19.2)
Caffeine	109 (13.7)
Benzocaine	95 (11.9)
Levamisole	53 (6.6)
Paracetamol	31 (3.9)
Procaine	28 (3.5)
Phenacetin	16 (2.0)
Dimethyl sulfone	11 (1.4)
Ketamine	7 (0.9)
Methamphetamine	3 (0.4)
Fluorexetamine/2‐fluoro‐2‐oxo PCE	2 (0.3)
Heroin	2 (0.3)
Dextromethorphan/methorphan/levomethorphan	1 (0.1)
Modafinil	1 (0.1)
N,N‐dimethylpentylone	1 (0.1)
N‐isopropylbenzylamine	1 (0.1)
Tetracaine	1 (0.1)
Tiletamine	1 (0.1)
Xylazine	1 (0.1)

Note: Samples with ‘/’ denote that FASS was unable to distinguish between drugs. Samples from both music festival and community seizures are included for samples where target drugs were identified as the main drug in a sample. The percentage represents the total number of samples where the specific adulterant was identified divided by the total number of samples where the target drug was identified as the main drug. Note that where the main drug is methamphetamine, MDMA, heroin, or cocaine, the adulterants are co‐detections with the main drug; where no main drug is identified, the adulterants could be the only compound present, or co‐detected with other drugs.

3.4. Novel Psychoactive Substances

Table 4 shows the number of samples where specific substances, including novel psychoactive substances, were detected; 133 samples (3.9%) had at least one of these substances detected. This analysis includes all samples from both music festival and community seizures. N,N‐dimethylpentylone (also known as dipentylone) was the most commonly detected substance. Novel benzodiazepines (bromazolam, clonazolam, deschloroetizolam, etizolam, flubromazepam and nimetazepam), anaesthetic/hallucinogens (fluorexetamine/2‐fluoro‐2‐oxo PCE and tiletamine) and novel opioids (acetylfentanyl and protonitazepyne/isotonitazepyne) were also detected.

TABLE 4.

Detections of specific substances, including new psychoactive substances, from music festival and community samples, New South Wales, April 2021 to March 2025.

Drug	Detections, n (%)
N,N‐dimethylpentylone	52 (1.5)
Fluorexetamine/2‐fluoro‐2‐oxo PCE	19 (0.6)
Tiletamine	16 (0.5)
1,4‐butanediol	7 (0.2)
Bromazolam	7 (0.2)
Chloromethcathinone	6 (0.2)
2C‐B	4 (0.1)
2‐fluoro‐2‐oxo PCE	4 (0.1)
Xylazine	3 (0.1)
Clonazolam	3 (0.1)
3‐hydroxy‐PCP	2 (0.1)
O‐acetylpsilocin	2 (0.1)
Etizolam	2 (0.1)
N‐cyclohexyl methylone	2 (0.1)
2‐methylmethcathinone	1 (< 0.1)
Acetylfentanyl	1 (< 0.1)
Deschloroetizolam	1 (< 0.1)
Eutylone	1 (< 0.1)
Flubromazepam	1 (< 0.1)
Fluorexetamine	1 (< 0.1)
Hydroxetamine	1 (< 0.1)
MDPBP	1 (< 0.1)
N‐cyclohexyl‐N‐methyl methylone	1 (< 0.1)
Nimetazepam	1 (< 0.1)
Protonitazepyne/isotonitazepyne	1 (< 0.1)

Note: Before 1 May 2024, 2‐fluoro‐2‐oxo PCE could not be distinguished from fluorexetamine. The ‘/’ denotes that FASS was unable to distinguish between drugs. The denominator is the total number of samples analysed (n = 3378). Where the number of NPS detections exceeds the sum across main drugs in Table 3, this indicates that the main drug was identified but not one of the target drugs.

The Supporting Information contains figures showing the combinations of drugs co‐detected within a sample.

3.5. Contribution to NSW Early Warning Surveillance System

NSW Health issued 33 drug warnings between January 2021 and March 2025; CoSMoSS data contributed information for 11 of these, highlighting its significance as a contributor to the NSW early warning drug surveillance system.

4. Discussion

This article describes the number, contents and purity of selected street‐level drug samples seized by NSW Police Force at music festivals and in the community located in selected police regions across NSW between January 2021 and March 2025, and analysed as part of the CoSMoSS program.

The purity of methamphetamine has been very stable over time (see Figure 2); median (free base) purity has been near 80% (equivalent to 99% hydrochloride salt purity) in every quarter monitored, and although some samples are low purity, the interquartile range of purities is small in every quarter. This continues a previously identified trend of high and stable methamphetamine purity in most Australian states since 2012–2013 [4]. Adulterants were uncommon in samples where the main drug was methamphetamine.

In contrast, samples where the main drug was cocaine showed considerable variation in purity and were commonly adulterated. However, the adulterants identified most commonly were usually of low potential to cause toxicity (e.g., caffeine or low purity lidocaine). NSW Health has issued safety information concerning higher doses of lidocaine in the past [10]; the quantification of lidocaine purity was instigated in CoSMoSS samples in response to this.

It should be noted, however, that NSW Health has issued multiple public drug warnings concerning adulteration of methamphetamine and cocaine with opioids, most commonly including heroin but occasionally nitazenes [14, 15] and that CoSMoSS detected two samples each of heroin in methamphetamine or cocaine. This suggests that despite the high stability of methamphetamine purity observed in CoSMoSS samples, continued monitoring of methamphetamine and cocaine contents and purity is worthwhile. The nitazene adulterants have been identified through a separate surveillance and early warning system in NSW, the Prescription, Recreational and Illicit Substance Evaluation program [16], which provides access to rapid and extensive toxicology and substance testing to NSW acute care services for cases with severe and unusual substance‐related toxicity or clusters of overdoses. This demonstrates that analysis of drug samples and clinical specimens associated with cases of unexpected overdose may yield complementary results, to be interpreted in the context of all available data surrounding drug markets in NSW.

Strong conclusions cannot be drawn about trends in purity of heroin over time, as street level heroin samples were scarce for sampling by the CoSMoSS program. Only a small proportion of samples where the main drug was heroin were found to contain an adulterant and the adulterants listed (caffeine and methamphetamine, see Table 3) were of low potential to cause toxicity. However, reductions in global heroin supply [17] may mean there will be increasing adulteration or substitution of heroin with other opioids which are cheaper and easier to manufacture [18], including novel opioids such as nitazenes and fentanyl analogues with higher potential to cause toxicity. Continued monitoring of adulterants in heroin (and other drugs) is critical to ensure that messaging about the potential to experience harm is communicated not only to people who use drugs, but also to health staff, police and chemical analysts.

MDMA was the most commonly detected drug, which likely partly reflects the CoSMoSS program's focus on street‐level seizures from music festivals. The proportion of festival samples where the main drug is MDMA was similar for CoSMoSS (61.5%) as for festival‐based voluntary drug checking services (e.g., at the five Victorian festival‐based drug checking services to November 2025, the percentage of samples containing MDMA ranged from 40% [19] to 63% [20]). The adulterants detected in samples with MDMA as the main drug were broadly similar to other reports across Australia [21].

Dose of MDMA tended to be higher in tablets than capsules, mirroring the result reported in the previous CoSMoSS publication [1]. Here, we provide additional information on dose and purity over time, for a longer period and at greater temporal detail than other Australian published data [3, 6]. We note the recent increase in the frequency of detections of tablets and capsules exceeding the 140 mg dose MDMA (free base) at which the risk of toxicity is substantially raised [11, 12, 13]. Further, the range of doses detected (from 5 to 243 mg) illustrates the risks of taking multiple MDMA capsules or tablets without prior knowledge of the dose.

Multiple drug alerts concerning high dose MDMA tablets have been issued in NSW in the period of interest; in general, alerts are not issued for high dose capsules in the absence of other findings to increase the chance of risk (e.g., adulterants) as the dose depends on the amount of crystalline substance or powder within the capsule, and this could be expected to vary considerably between capsules. Further, capsules are less distinctive than tablets in colour, shape and markings, and thus issuing targeted public health messaging is not straightforward. The finding of high dose MDMA tablets and capsules remains an important and actionable public health message, and the decision to release information is supported across jurisdictions in Australia. Additionally, dangerous substances including nitazenes have been detected in tablets with colour and markings typical of MDMA [22], which is important to consider when crafting public health messaging. Finally, tablet colour and appearance alone are not reliable indicators of contents, purity or dose, which also complicates public health messaging.

The present methodology differs from analyses of data from large seizures [4] in its focus on small street‐level seizures. The benefit of this is that it provides a more direct picture of the contents and purity of drugs at the level of consumption. It also differs from front‐of‐house drug checking (where samples are voluntarily submitted for analysis prior to planned use). In that setting, the submitter clarifies what the substance is believed to be, and it therefore can be inferred that any substances identified in samples found to contain other substances were in fact misrepresented to consumers. For samples tested in CoSMoSS, we cannot make this same inference, as no information is collected about what the consumer expects their sample to contain.

Most of the samples are found to contain one of the four target drugs targeted by design, with the resulting weakness that relatively few samples of other drugs have been selected, including ketamine and gamma‐hydroxy butyrate (GHB), which have each shown increasing prevalence of use in national surveys [23], as well as increasing harms [24]. In the later part of this study period, a decision was made to sample more ketamine where available, but sampling is still considerably lower than other drugs. Nevertheless, CoSMoSS has contributed valuable information to the NSW early warning system, and further expansion could be considered (e.g., sampling from all police regions and more street‐level drug types) to complement the existing program and other surveillance sources, depending on system resources.

5. Conclusions

Analysis of seized substances provides high quality objective information about contents and purity of illicit drugs in circulation at street level. This information, along with harm reduction messaging, can be disseminated to the community when high risk/adulterated substances are identified. The results of this study can also inform the public about the usual contents, purity and dose of drugs suspected by police to be MDMA, heroin, cocaine or methamphetamine and later, ketamine. This program continues to provide valuable and timely intelligence to the early warning surveillance system in NSW, and highlights the feasibility and value of a strong partnership between laboratories, police, public health organisations, clinicians and communities.

Author Contributions

All authors made a substantial contribution to the intellectual content and preparation of the article, and all authors take public responsibility for the content of the article. All authors have given final approval before publication.

Funding

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Count and distribution of (free base) purity (%) by year (by date of seizure) and by main drug detected in sample (and form, for MDMA), from January 2021 to March 2025.

Table S1: Pairwise adjusted p values comparing median purity between years by drug and form, for drug forms where the omnibus test was significant.

Figure S2: Count and distribution of average weight per unit of MDMA tablets and capsule contents, by year and quarter (by date of seizure) and form of MDMA, from January 2021 to March 2025.

Figure S3: Count and distribution of average MDMA dose (free base) by year of seizure and form of MDMA from January 2021 to March 2025.

Table S2: Pairwise adjusted p values comparing median weight and dose between years for MDMA capsules.

Figure S4: Co‐detections where the main drug was cocaine (Note: samples with only cocaine detected are not shown here).

Figure S5: Co‐detections where MDMA was the main drug.

Figure S6: Co‐detections where methamphetamine was the main drug.

Figure S7: Co‐detections where the main drug was ‘none identified’.

Data Availability Statement

Deidentified data that support the findings of this study may be made available to investigators whose proposed use of the data is approved by the NSW Ministry of Health and/or NSW Police Force, on request from the corresponding author of the publication.