Clinical trials with cocaine-dependent outpatients have found a strong relation between in-treatment and follow-up abstinence, and the strength of this relation is constant across treatment conditions with variable efficacy in generating abstinence. The authors conducted secondary analyses of data from 3 clinical trials to determine whether this relation generalizes to cocaine-dependent homeless persons. The 3 trials (total N = 543) were conducted in a community health care facility for homeless people. The 7 treatment arms across the 3 trials were combinations of day treatment, abstinence-contingent housing, and vocational training. Drug use was measured with urine toxicology testing. Consecutive weeks of abstinence during treatment were strongly related to abstinence at the 12-month follow-up, whether or not missing 12-month data were included in the analysis. The treatment arms differed in their efficacy in generating abstinence, but the relation between in-treatment and follow-up abstinence did not differ across treatment arms. These results replicate earlier reports of these relations and extend them to a population of homeless people. The lack of differences between treatment arms in the in-treatment-follow-up abstinence relation implies that that relation is independent of the treatment-specific intervention components that generate group differences in abstinence. © 2009 American Psychological Association.