The impact of length of stay in drug abuse treatment on follow-up drug use and criminal behavior has important clinical and policy implications. In this paper we use longitudinal data from the Treatment Outcome Prospective Study to estimate the simultaneous effects of time in treatment and employment outcomes--weeks worked and total earnings--on posttreatment drug use and criminal activity. The drug use/criminal activity variables include four indexes measuring the severity of use, drug-related problems, predatory illegal acts, and overall criminal behavior. The results show that time in treatment had a negative and statistically significant impact on these outcome variables for every modality with residential clients experiencing the largest relative impact. The time-in-treatment effect was robust even when employment outcomes were modeled jointly with drug use and criminal activity outcomes. These findings highlight the importance of length of stay in treatment in the recovery and rehabilitation of drug abusers.