Objectives The benefit of the additional use of AD in the inpatient treatment of depression with intensive cognitive-behavioral therapy (CBT) has been investigated in a naturalistic design. Methods Depressiveness was assessed using the Beck Depression Inventory (BDI-II) during a preliminary interview (T0), at admission (T1), at discharge (T2), and at a 6-month follow-up (T3). Two study phases were compared: During Phase A, AD were recommended in accordance with the German guideline. In Phase B, AD were no longer recommended, and they were only prescribed upon explicit request from patients. In phase A (N = 574), 60.3% of all patients were taking AD at discharge. In Phase B (N = 424), 27.9% of patients were on AD at discharge. Apart from the difference in AD usage, the two treatment conditions were similar, and the samples did not significantly differ in terms of age, sex, diagnoses, history of suicide attempts, comorbid anxiety disorders, and unemployment. Results In both study phases, BDI-II scores were strongly decreased at T2 and T3, respectively, compared with T1. The BDI-II scores of the two phases did not differ at any of the measurement time points. Depression changes were similar in both phases. In sequential multiple regression analyses with the total sample, AD were no significant predictors for the reduction of depression at either T2 or T3. Conclusions The inpatient CBT was effective in depression. The effectiveness of CBT is not improved by the additional use of AD. The current prescribing practices of AD should be questioned.