BACKGROUND There are few published placebo-controlled clinical trials demonstrating the efficacy of the newer antidepressants in markedly or severely depressed hospitalized patients. This study demonstrates the efficacy of nefazodone compared with placebo in the treatment of patients hospitalized for major depression. METHODS Nefazodone and placebo treatment were compared in a 6-week trial of 120 patients hospitalized for DSM-III-R diagnosed major depression (without psychosis) at 2 study centers. Efficacy was evaluated using standard psychiatric rating scales, and patients were monitored for safety. RESULTS Nefazodone treatment resulted in a significant reduction (p < .01) of the 17-item Hamilton Rating Scale for Depression (HAM-D-17) total score compared with placebo from the end of the first treatment week through the end of the study (-12.2 nefazodone vs. -7.7 placebo). At the end of the trial, significantly more nefazodone-treated patients (50%) than placebo-treated patients (29%) had responded, as indicated by their Clinical Global Impressions-Improvement score (p = .021) or by a > or = 50% reduction in their HAM-D-17 scores (p = .017). Significantly more patients treated with nefazodone (36%) than placebo-treated patients (14%) had a HAM-D-17 score < or = 10 at the end of treatment (p = .004). Significant treatment differences (p < .01) in favor of nefazodone were also seen in the Montgomery-Asberg Depression Rating Scale; the HAM-D retardation, anxiety, and sleep disturbance factors; and HAM-D item 1 (depressed mood). Patients with dysthymia in addition to major depression also showed significant improvement (p < .05) when treated with nefazodone, with significant differences in response rates seen as early as week 2 and through the end of the trial. The mean nefazodone dose was 491 mg/day at the end of week 2 and 503 mg/day at the end of treatment. Nefazodone was well tolerated, and the number of patients discontinuing owing to adverse events was small, with no significant safety issues noted in either treatment group. Fewer nefazodone-treated than placebo-treated patients discontinued owing to lack of efficacy. CONCLUSIONS Nefazodone was superior to placebo in the treatment of marked to severe major depression in patients requiring hospitalization. The clinical benefit of nefazodone was evident as early as the first week of treatment as judged by several measures of efficacy, with significant differences from placebo sustained throughout the trial.