Sleep/waking and body temperature (Tb) were recorded in male rats in a 12:12 light-dark photoperiod at one of 3 ambient temperatures (Ta's): 20, 30, or 32 degrees C. After adaptation to the sleep recording chamber for at least 48 h, the rats were injected with saline at the beginning of lights-on (day S1). Twenty-four hours later (day P1), they were injected with PCPA (300 mg/kg, i.p.) and recordings continued for 4 more days (P2-P5). At these Ta's, hypothalamic 5-HT was depleted by 66-75% 30 h post-PCPA. Changes in both amplitude and acrophase of Tb depended on Ta. Compared to S1, amplitude was reduced on P2-P4 at 20 degrees C and on P3-P4 at 30 degrees C. Acrophase was advanced on P1-P3 at Ta 20 degrees C only. Sleep variables were generally independent of Ta and largely unchanged in the dark. In the light, amounts of slow-wave sleep (SWS) were depressed on P2-P4, number of bouts decreased on P3-P5 and percent nocturnality decreased on P2-P5. Bout length was depressed on P2 and lengthened on P4-P5. Acrophase was delayed on P2-P4 at Ta 30 degrees C. Amounts of rapid-eye-movement sleep (REMS) were depressed on P1-P3. REMS bout length decreased on P1-P3. The decreases in number of REMS bouts seen on P1-P3 depended on Ta. Changes in percent nocturnality and acrophase of REMS were minor. Waking----SWS transitions decreased on P3-P5 while SWS----REMS transitions were reduced on P1-P2. These results suggest that PCPA affects circadian aspects of both Tb and sleep, that 5-HT is important in the initiation of SWS bouts, and finally that the mechanisms by which 5-HT depletion affects Tb, SWS and REMS are different.