Cocaine dependent methadone patients were randomly assigned to 6 months of high intensity cognitive-behavioral therapy or low intensity therapy. A repeated measures ANOVA was conducted with patients stratified on severity of cocaine use at baseline. Both treatment groups showed significant and equivalent reductions in cocaine use during the post-treatment period. Completing either therapy and lower cocaine severity at baseline were associated with lower proportion of cocaine-positive urines across a 48-week post-treatment period. Examination of the treatment x cocaine severity interaction provided some evidence that high-severity patients improved more if exposed to high intensity treatment than to low intensity treatment. Positive outcomes for therapy completers relative to non-completers increased over time. The results are consistent with several clinical trials showing that: (1) participation in treatment is associated with reductions in cocaine use; and (2) the relationship between treatment intensity and outcome is not linear and may better be explained by an interaction between patient and treatment factors.