Electroencephalographic sleep was quantitated in adult male Sprague-Dawley rats following single injections of the methylesters of tryptophan, valine or alanine. The amino acids were administered at the onset of the daily light period (09.00 h); electrographic data were collected for the succeeding 6-h period. Saline served as the injection control, and fluoxetine, a serotonin-reuptake blocker, as a positive control. The injection of tryptophan methylester (125 mg/kg) caused a delay in rapid eye movement (REM) sleep onset, and significantly reduced the amount of REM sleep during the first 2 h postinjection. Tryptophan produced no effect on sleep onset, nor did it influence total sleep time. Fluoxetine (2.5 mg/kg) produced similar effects, as previously observed. The methylesters of valine and alanine were without effect on REM sleep, when injected at a molar dose equivalent to that for tryptophan. No consistent effects of any of the test substances were noted on non-REM (NREM) sleep or waking time, or on any of the other sleep indices quantitated. Together, the data indicate that tryptophan selectively reduces REM sleep; the effect is not due to a non-specific action of amino acids or their methylesters. The effect on REM sleep may be the consequence of a tryptophan-induced stimulation of 5-HT synthesis and release, since it is like that produced by fluoxetine, a drug that enhances transmission across serotonin synapses.