This study compared homeless mentally ill chemical abuser (MICA) clients (n
= 342), male and female, sequentially assigned to either of two modified t
herapeutic community programs (TC1 and TC2) and to a treatment-as-usual (TA
U) control group. Follow-up interviews were obtained at 12 months postbasel
ine and at time F (on average more than 2 years postbaseline) on a retrieve
d sample df 232 (68%) clients and 281 (82%) clients; respectively. Outcome
measures assessed five domains: drug use, crime, HIV risk behavior, psychol
ogical symptoms, and employment. Individuals in both modified TC groups sho
wed significantly greater behavioral improvement than TAU at 12 months and
time F, and the modified TC2, with lower demands and more staff guidance, w
as superior to modified TC1. Completers of both TC programs showed signific
antly greater improvement than dropouts and a subgroup of TAU clients with
high exposure (i.e., more than 8 months) to other treatment protocols. The
present findings support the effectiveness and longer term stability of eff
ects of a modified TC program for treating homeless MICA clients.