This randomized controlled trial examined whether an abbreviated treatment using abstinence contingency management for housing and work/training (CM, n=103) with cocaine dependent, non-psychotic, dually diagnosed homeless persons, would show non-inferior outcomes compared to the full treatment (CM+, n=103). It was hypothesized that CM+ would show superior abstinence and retention outcomes, but that CM, with components derived from previously effective behavioral day treatment, would obtain non-inferior outcomes, defined as 75% or more of those observed for the full treatment (CM+) during active treatment months 1-6. The CM+ included all CM components but added behavioral day treatment that included voucher reinforcement of $7.50-15.00, for objective weekly therapeutic goal attainment in five domains of functioning: drug dependence, homelessness, unemployment, non-drug related recreational activities, and behavioral, health, or mental health problems. Abstinence was assessed by observed urine specimen collection and weekly testing weeks 1-52, and randomly, bi-monthly for months 13-18. Abstinence, homelessness, employment and other outcomes were also assessed at baseline, 2, 6, 12, and 18 months.

Inclusion Criteria: McKinney Act criteria for homelessness; DSM IV diagnosis for Cocaine Dependence; self-reported use within the last 2 weeks; psychological distress indicated by score of 70 on 1 or more Brief Symptom Inventory scales; willingness to participate; no plans to move from Birmingham area for 18 months;

-

Exclusion Criteria: cognitive impariment precluding informed consent; requiring immediate inpatient medical treatment