The acute and chronic effects of alcohol and alcohol withdrawal on sleep patterns and plasma GH and cortisol fluctuations occurring during sleep were studied. Five healthy men, aged 21-26 yr, consumed a placebo drink for 3 baseline nights, alcohol (0.8 g/kg) for 9 nights, and a placebo drink on final withdrawal night. All-night polygraphic sleep recordings and blood samples (every 20 min with a venous catheter) were collected for 1 placebo, 1 acute alcohol, 1 chronic alcohol (night 9), and 1 alcohol withdrawal night. Acute and chronic alcohol consumption reduced rapid eye movement sleep nonsignificantly during the first half of the night, whereas slow wave sleep (stages 3 and 4) was increased significantly after acute alcohol, returning to baseline values on the chronic alcohol night. On the withdrawal night, rapid eye movement sleep and slow wave sleep were not significantly different from placebo sleep. Alcohol significantly suppressed plasma GH values (70-75%) on acute and chronic nights. All measures of GH, including total integral for bedtime hours, mean hourly rate, and peak level, were similarly affected by alcohol. GH returned to placebo values on the withdrawal night. Measures of nighttime plasma cortisol were not significantly altered by alcohol or alcohol withdrawal at this dose level.