We hypothesized that stimulant abusers would sleep more and have more rapid eye movement (REM) sleep than primary alcoholics during acute withdrawal (first 10 days drug free) but would have opposite patterns during subacute withdrawal (days 11-14 drug free). We compared polygraphic sleep patterns during acute withdrawal (days 3-10) for 7 stimulant abusers and 8 alcoholics and during subacute withdrawal (days 11-14) for 7 different stimulant abusers and 8 different alcoholics. For comparison purposes, a group of normal controls from our preexisting database were matched for age and gender. Two statistically significant interactions were found: consistent with our hypothesis, stimulant abusers showed greater total sleep (TST) and REM sleep during acute withdrawal than subacute withdrawal, compared with alcoholics. In contrast, alcoholics showed less TST and REM sleep during acute withdrawal than during subacute withdrawal. Our polygraphic sleep data support the hypothesis that physiological withdrawal differs in alcoholics compared with stimulant abusers. Different mechanisms may underlie withdrawal in these two substances.