The efficacy of nefazodone in prevention of relapse of depression was evaluated in a 36-week double-blind, placebo-substitution, continuation treatment trial. After 16 weeks of acute, single-blind treatment with nefazodone, 131 patients responding to treatment and in stable remission were randomized in a 36-week double-blind trial to either nefazodone (n = 65) or placebo (n = 66). Patients were defined as having relapsed if they had a total score > or = 18 on the 17-item Hamilton Depression Scale on two consecutive visits or if they discontinued treatment for lack of efficacy. Relapse rates were significantly lower for patients randomized to continued nefazodone treatment than for patients switched to placebo. Kaplan-Meier estimates of relapse rates 9 months (36 weeks) after the end of acute treatment were 1.8% for nefazodone versus 18.3% for placebo (P = 0.009) by the Hamilton Depression Scale and 17.3% versus 32.8% (P = 0.028) by discontinuation for lack of efficacy. The mean modal dose of nefazodone was 412 mg/day at study endpoint. These results demonstrate the clinical effectiveness of up to 1 year's treatment (16 weeks acute and 36 weeks continuation) with nefazodone in depressed patients. Long-term efficacy of nefazodone was accompanied by a good safety profile without any weight gain and with minimal symptoms of withdrawal upon abrupt discontinuation of treatment.