Ethanol is known to alter central neurotransmission and endocrine functions. Urine melatonin was studied in 10 male chronic alcoholic patients, before and after two weeks of controlled alcohol abstinence, and in sex and age matched healthy controls. In both groups, 24-hour urines were collected in two fractions corresponding to day- (D) (08:00-20:00) and night- (N) (20:00-08:00) time. Urine melatonin was assayed by RIA after methylene chloride extraction. Twenty-four hour urine melatonin levels were calculated adding up D and N values. In patients during alcohol intake, the 24-hour urine melatonin levels were significantly higher than in controls (p = 0.004, Student's t test). A disruption of the physiological ratio between N and D values was also observed, since the higher melatonin levels occurred in the D fraction. In drinking alcoholics, melatonin D values were significantly higher than the D values found in controls (p less than 0.01, Student's t test) and in the same patients after alcohol withdrawal (p less than 0.05). The N/D ratio approximated 1 during alcohol intake and became larger than 1 after alcohol withdrawal, as in the controls. The melatonin data were correlated with the suppressive effects of dexamethasone (DXT) on cortisol secretion evaluated both during alcohol intake and during abstinence. After alcohol withdrawal, the two (out of 10) patients, who remained unresponsive to the DXT suppression test, showed high D melatonin values and a low N/D ratio. These preliminary data indicate that in chronic alcoholism the pattern of urinary "melatonin- like immunoreactivity" is altered.