Methods	Design: Randomized control trial
	Recruitment: Clients were recruited into Project H&ART by staff of a day shelter for homeless persons, outreach or clinic staff of Albuquerque Health Care for the Homeless (HCH), or by one of the community agencies which provides other services to homeless persons.
	Setting: Housing programs and motel‐like accommodations in Albuquerque, New Mexico, US
	Follow‐up: 10 months
Participants	Population: Homeless alcohol abusers
	Eligibility criteria: Individuals had to be homeless, single adult alcohol abusers who had been in the Albuquerque area for at least three months.
	Sample size: Total n = 469; Group‐1 n = 161, Group‐2 n = 164, Group‐3 n = 92, Group‐4 n = 52.
	Baseline characteristics: “Clients ranged in age from 18‐67, with a median age of 37 years. Baseline comparisons among persons in the four intervention groups revealed no differences in age group, years of education, race/ethnicity, or classification as having alcohol, drug, housing stability, employment, or legal problems. In each of the groups over 85% of the clients reported alcohol as their primary substance of abuse. The majority of the clients who entered Project H&ART were males; females represented only 13% of the client population. About 41% were non‐Hispanic white (referred to subsequently as “white”); 31% Hispanic white (Hispanic); 18% Native American; and 10% belonged to other race groups. There were some differences in demographic characteristics among members of the different race/ethnic and gender groups. Whites had somewhat higher education levels, with about one third of the population having completed more than 12 years of school. Women were significantly less likely than men to be veterans”
Interventions	Group 1, the high intensity group, received case management and substance abuse counselling services, along with four months of housing in four‐plex apartment buildings staffed by residence managers who provided peer support.
	Group 2, medium intensity group, received four months of housing in similar apartments with support services from peer residence managers. Clients in Group 2 were expected to seek treatment for their alcohol and drug abuse on their own initiatives, from services normally available in the community.
	Group 3, low intensity group, received four months of apartment‐ or motel‐based housing and no additional services.
	***About halfway through the 16‐month intervention phase, Group 3 housing services were discontinued due to safety concerns for staff and clients. Individuals randomized to the new low intensity nonhoused group (designated Group 4) received referrals and bus fare to local and statewide alcohol treatment agencies and were paid to provide health services utilization data at twice weekly check‐ins.
	*In all three groups subjects were required to be abstinent from substances of abuse and were subjected to random, and “on demand” breath and urine testing. Those who could not maintain sobriety were discharged from the program.
Outcomes	Housing stability: Percentage of participants who had stable housing was measured using the Personal History Form (PHF).
	Substance use: the number of days of alcohol use using the Addiction Severity Index (ASI).
	Employment: Number of days of employment was measured using the Personal History Form.
Notes	The only exception to the randomization process was that after two women were randomly assigned to the nonhoused control group a decision was made to randomize all subsequent female participants to one of the housed groups