This study was designed to test the effect of dietary protein on blood levels of insulin and glucagon. Twelve normocholesterolemic (less than 200 mg/dl) and 11 hypercholesterolemic greater than 240 mg/dl) healthy male subjects, 31-62 years of age, were randomly given 3 liquid test meals 1 week apart. Meals were identical except for the protein source (soybean, casein, or protein free). Blood was drawn at fasting, and 0.5 and 2 h postprandially. Insulin and glucagon levels were measured by radioimmunoassay. Hypercholesterolemic subjects had a higher (P less than 0.05) insulin/glucagon ratio (1.5) than normocholesterolemic subjects (0.7) 2 h post-prandially when fed the casein test meal. There was no significant difference following the soybean test meal. This implies that the post-prandial insulin/glucagon ratio was affected by the amino acid composition of the diet. There was a consistently higher insulin response to all test meals among hyper- versus normocholesterolemic subjects. These results are consistent with our hypothesis that the hypocholesterolemic effects of soybean protein and the hypercholesterolemic effects of casein were mediated by altered levels of insulin and glucagon.