Objective: We examined comorbid disorders' prevalence, their impact on abstinence, and the impact of depressive symptoms on abstinence and of abstinence on depressive symptoms. Method: A randomized controlled trial's data on outcomes from treating cocaine dependence were used. It compared abstinencecontingent housing and work to contingency management plus behavioral day treatment. Regardless of original trial arm assignment, groups of participants with no additional Axis I disorders (n=87) and 1 or more additional Axis I disorders (n = 113) were compared for abstinence. Changes in depression symptoms, measured by the Beck Depression Inventory, were analyzed as a function of 4 cohorts of increased consecutive weeks abstinent. An autoregressive cross-lagged path model examined reciprocal relationships between depression and abstinence. Results: Most prevalent additional disorders were depressive disorders, followed by anxiety disorders. Additional disorders did not significantly affect abstinence. Cohorts with more abstinence were linearly related to lower depression symptoms. The cross-lagged model showed that longer abstinence predicted decreases in depressive symptoms at 6 months. However, depressive symptoms did not predict changes in abstinence. Conclusions: Our study adds to others that have found an effective treatment targeted at specific problems such as substance abuse, social anxiety disorder, and posttraumatic stress disorder that may have the side benefit of reducing depression. Additionally, we find that depression does not interfere with effective substance abuse treatment for cocaine dependency. This may be the 1st formal analysis comparing the ability of cocaine abstinence to predict future depressive symptoms versus depressive symptoms to predict future cocaine abstinence.