Recent interest in increasing access to substance abuse treatment for drug-involved offenders has been spurred by concerns over expanding prison and jail populations, high recidivism rates for drug-involved offenders, and the close link between illegal drug use and criminal activity. Chronic untreated drug and alcohol abuse is likely to result in high rates of repeated contacts with the criminal justice system and a greater likelihood of reincarceration. Unless these offenders naturally desist from drug use, or are successfully engaged in treatment, recidivism is likely to remain high and the courts and correctional systems are likely to continue to be overwhelmed by large numbers of drug-involved offenders. This article uses multiple recidivism measures to assess the long-term effects of diversion to a highly coercive, long-term residential therapeutic community treatment for repeat felony drug offenders charged with drug sales and facing mandatory incarceration in state prison. If the offenders completed the 18-24 month program, all charges were dismissed; dropouts were returned to court for prosecution and sentenced to state prison. Compared with a closely matched sample of offenders sentenced to prison, Drug Treatment Alternative to Prison (DTAP) program participation generally reduced the prevalence and annual rate (adjusting for time in the community) of recidivism, and delayed time to first rearrest. In multivariate models of rearrest prevalence and adjusted annual rearrest rate, DTAP program participation was related to lower recidivism at significance levels between 0.05 and 0.10, after controlling for criminal history and other covariates. These findings suggest that diverting high-risk, prison-bound felony drug sellers to long-term treatment can yield significant, long-term reductions in recidivism. (Contains 6 tables, 1 figure, and 1 note.)