This study sought to assess the efficacy of treatment for cocaine abuse and to compare the relative effectiveness of a cognitive-behavioral relapse prevention treatment with that of a Twelve-Step recovery support group in an outpatient group treatment setting. One hundred ten subjects seeking treatment were alternately assigned to relapse prevention or Twelve-Step treatment. Self-report data were collected at baseline, posttreatment, and 6-month follow-up. There were no differentia] effects of treatment type on cocaine or marijuana use over time. However, subjects in both treatment conditions reduced cocaine and marijuana use at post-treatment. Subjects in both groups reduced their alcohol use from pretreatment to posttreatment. Subjects receiving Twelve-Step treatment showed greater increases from posttreatment to 6-month follow-up in alcohol use than did relapse prevention participants. Treatment attendance was negatively related to cocaine use at posttreatment and cocaine and marijuana use at 6-month follow-up. Difficulties in conducting cocaine treatment outcome research are discussed as are treatment and research implications of the findings.