We studied the effect of acute (1 day) and subacute (16 days) administration of the new antidepressant, nefazodone (400 mg daily), and the selective serotonin re-uptake inhibitor (SSRI), paroxetine (30 mg daily), on the sleep polysomnogram of 37 healthy volunteers using a random allocation, double-blind, placebo-controlled design. Compared to placebo, paroxetine lowered rapid eye movement (REM) sleep and increased REM latency. In addition, paroxetine increased awakenings and reduced Actual Sleep Time and Sleep Efficiency. In contrast, nefazodone did not alter REM sleep and had little effect on measures of sleep continuity. We conclude that in contrast to typical SSRIs, nefazodone administration has little effect on sleep architecture in healthy volunteers.