This study examined posttreatment patterns of polysubstance use and heroin relapse in a sample of 43 adolescents (ages 14-20) entering short-term residential treatment for primary heroin use. At 12-month follow-up, youths that achieved heroin abstinence (N = 19) were significantly less likely than youths that relapsed to heroin (N = 24) to endorse polysubstance use and cannabis, cocaine, or benzodiazepine use. Furthermore, heroin-abstinent youths significantly reduced their cannabis and cocaine use across the study period while youths that relapsed made initial reductions before returning to their pretreatment levels of use for these drugs. Clinical implications for heroin-using youths and areas for future research are discussed. (Contains 2 figures, 2 tables and 2 notes.)