In previous uncontrolled studies, nefazodone, a new antidepressant drug, increased REM sleep or had no effect on REM sleep. We report a double-blind, placebo-controlled, parallel group study on effects of nefazodone (N) on polysomnographic sleep variables in healthy human volunteers. Nefazodone was administered for 16 consecutive days, and nocturnal sleep, as well as multiple sleep latency test (MSLT), was monitored before, during, and after N administration. We found that N had no effect on any measured REM sleep variable including REM sleep duration, REM density, and nocturnal REM sleep distribution. Nefazodone also had no significant effect on nocturnal total sleep time or NREM variables, but increased daytime alertness measured by the MSLT. Since N is a potent serotonin reuptake blocker, the present findings that N had no effect on REM sleep cast doubt on the hypothesis that antidepressant drugs decrease REM sleep by increasing serotonergic neurotransmission. A review of other relevant work also casts doubt on this hypothesis.