To determine whether ethanol inhibits nocturnal melatonin (MT) secretion, three experiments (A, B, and C) were performed in seven normal subjects. In A, ethanol at a dose of 0.34 g/kg was administered orally at 6:00, 8:00, and 10:00 PM. Each dose was increased to 0.52 g/kg in B. In C, water was substituted for ethanol. Blood samples for determination of serum MT levels were drawn every second hour between 6:00 PM and 8:00 AM. Urinary excretion of MT during the night was also determined. In A, serum ethanol reached a maximal level of 13 +/- 1 mmol/L at 12 midnight. In B, the corresponding maximum was 25 +/- 1 mmol/L. The higher alcohol dose inhibited nocturnal MT secretion by 20% +/- 5% (P < .01), whereas the lower dose lacked such effect. Urinary excretion of MT was left unaffected by alcohol at both doses. Five additional normal subjects were given alcohol as described above at a dose of 0.52 g/kg (experiment D). This induced mild nocturnal hypoglycemia as evidenced by a glucose decremental area (5.9 +/- 1.8 mmol/L.h) that differed significantly from zero (P < .05). To determine whether a reduced glucose delivery to pinealocytes might contribute to the decreased MT secretion in alcohol-intoxicated subjects, two experiments (E and F) were performed in eight healthy individuals. In E, ethanol was given orally as in B; three small oral doses of glucose were also given at 8:00 PM, 10:00 PM, and 12 midnight. In F, water was substituted for ethanol and glucose.(ABSTRACT TRUNCATED AT 250 WORDS)