Oral glucose tolerance tests were conducted in 10 noninsulin-dependent diabetic and 14 healthy control subjects with a 75-g glucose load. The tests were repeated 1 week later with 43 g of ethanol mixed with the glucose. Blood samples were analyzed for ethanol, glucose, insulin, C-peptide, and glucagon levels. The blood ethanol peak was nearly equal in diabetic and control subjects (mean +/- SEM values of 55 +/- 8 and 48 +/- 6 mg/dl 45 min after ethanol ingestion). Ethanol did not affect glucose tolerance in either of the study groups. Mean +/- SEM values of the sum of the increment above the baseline glucose level were 659 +/- 48 vs. 675 +/- 76 mg/dl with or without ethanol in diabetics and 227 +/- 35 vs. 244 +/- 36 mg/dl in control subjects. The plasma insulin and C-peptide responses to glucose were delayed in diabetic patients compared to controls but were not affected by ethanol. In vitro, ethanol, at a concentration of 100 mg/dl or greater, significantly decreased insulin binding to erythrocytes in a dose-related manner. Scatchard analysis of competitive insulin binding to erythrocytes indicated that ethanol reduced insulin binding affinity (1.6 +/- 0.5 vs. 4.2 +/- 0.8 x 10(8)/M), but not binding capacity (4.5 +/- 2.4 vs. 4.4 +/- 1.7 nM, with and without ethanol, respectively).