Contingency management (CM) interventions usually use vouchers as reinforcers, but a new technique awards chances of winning prizes. This study compares these approaches. In community treatment centers, 142 cocaine- or heroin-dependent outpatients were randomly assigned to standard treatment (ST), ST with vouchers, or ST with prizes for 12 weeks. CM patients remained in treatment longer and achieved greater durations of objectively confirmed abstinence than did ST patients; CM conditions did not differ significantly. Although abstinence at 6- and 9-month follow-ups did not differ by group, the best predictor of abstinence was longest duration of abstinence achieved during treatment. Thus, prize and voucher CM systems are equally efficacious in promoting long durations of abstinence, which in turn are associated with benefits posttreatment.