After 4 adaptation nights, IM doses of morphine sulfate (7.5, 15, 30 rag/70 kg) and placebo were studied in 8 male postaddicts, using a cross-over design with randomized block analysis to demonstrate significant drug effect. Morphine decreased the number and duration of REM periods, delayed the onset of the first REM period, and possibly increased the time between REM period onsets. Thus, morphine significantly decreased REM state (REMS), but did not alter the pattern of maximum REMS in late night. Morphine significantly increased waking state (WS) but did not alter the pattern of maximum WS in early night. Morphine significantly increased tension, and shifted maximum tension from early night to middle night. Morphine altered EEG sleep patterns; although NREM definitions were thus less certain, morphine appeared to increase NREM light sleep (stages 1 and 2) and decrease NREM deep sleep (stages 3 and 4). After an initial decrease in REMS by 30 rag/70 kg of morphine, an increase in REMS was noted in the third night of the 2 pilot study subjects. The arousal response to morphine seen in postaddicts needs further study in other populations and during the course of chronic morphine use.