Two versions of cognitive-behavioral therapy (CBT), one with religious content (RCT) and one with standard protocol (NRCT), were used to treat 19-20 religious patients each. Fifty-nine religious patients who met the Research Diagnostic Criteria for nonpsychotic, nonbipolar depression were treated in 18-20 1-hr sessions over 3 months. Religious and nonreligious therapists were used in each CBT group. Pastoral counseling (PCT) treatment-as-usual and waiting-list control (WLC) conditions each contained 10-11 patients. RCT and PCT patients reported significantly lower posttreatment depression and adjustment scores than did either the NRCT or the WLC condition. The CBT difference was due largely to superior performance of the nonreligious therapists (with dissimilar values to the patients) in the RCT over the NRCT condition. Improvement in the three treatment conditions was equal at 3-month and 2-year follow-ups and greater than posttreatment WLC improvement levels.