This study evaluated the efficacy of cognitive-behavioral therapy (CBT) and 12-step facilitation (12SF) in treating cocaine abuse. Participants (N = 128) were randomly assigned to treatment conditions and assessed at baseline and at Weeks 4, 8, 12, and 26. Treatment lasted for 12 weeks. It was hypothesized that participants treated with CBT would be significantly more likely to achieve abstinence from cocaine than participants treated with 12SF. A series of patient-treatment matching hypotheses was also proposed. Across 2 different outcome variables, it was found that participants in CBT were significantly more likely to achieve abstinence than participants in 12SF. In addition, some support for matching hypotheses was found, suggesting that both psychotherapies may be differentially effective for identified subgroups of persons that abuse cocaine.