Marijuana extract, given in daily doses containing 70 to 210 mg delta-9-tetrahydrocannabinol (THC), induced effects on sleep that were virtually identical to those produced by the same doses of relatively pure (96%) THC. Both drugs reduced eye movements density with some tolerance developing to this effect. Stage 4 tendend to increase with drug administration. Abrupt withdrawal led to extremely high densities of eye movement, increased rapid eye movement (REM) durations, and a sharp but transient fall in stage 4 to baseline levels. These effects may be useful in the elucidation of the pharmacology of sleep. The effects on sleep of THC administration (but not withdrawal) closely resemble those induced by lithium. For this reason, we suggest further studies of THC in affective disorders. Evidence available thus far suggests that THC produces dysphoric symptoms in unipolar but not in bipolar depressed patients; these differences in response may prove of diagnostic value. An adequate therapeutic trial of THC in bipolar depressed patients has not yet been carried out.