Ninety-eight male cocaine-dependent patients who completed an intensive outpatient program (IOP) were randomly assigned to either standard group counseling (STND) or individualized relapse prevention (RP) aftercare. Heavier cocaine and alcohol use during IOP and low self-efficacy predicted more cocaine use during the treatment phase of the study, whereas lifetime diagnoses of alcohol dependence, major depression, and any anxiety disorder predicted less cocaine use. Rates of complete abstinence during the 6-month study period were higher in STND than RP, whereas RP was more effective in limiting the extent of cocaine use in those who used during Months 1-3. Matching analyses indicated patients who failed to achieve remission from cocaine dependence during IOP and those with a commitment to absolute abstinence did better in RP than in STND, whereas patients with other abstinence goals did better in STND than RP. Several differences in experiences before cocaine use and "near-miss" episodes were also identified.