To investigate the influence of dietary manipulation on in vivo glucose metabolism, we pair fed normal rats for 10 days with one of three diets: 1) high protein-low carbohydrate (Hi-PN) (n = 20); 2) intermediate protein (I-PN) (n = 11); and 3) low protein-high carbohydrate (Lo-PN) (n = 18). Fasting glucose, postmeal plasma glucose, and insulin concentrations were as follows: 118 +/- 2 mg/dl, 138 +/- 2 mg/dl, and 4.0 +/- 0.2 ng/ml in Hi-PN; 111 +/- 3 mg/dl, 147 +/- 3 mg/dl, and 5.1 +/- 0.3 ng/ml in I-PN; 102 +/- 2 mg/dl, 162 +/- 2 mg/dl, and 6.0 +/- 0.2 ng/ml in Lo-PN, respectively. Basic hepatic glucose production (HGP) was 6.6 +/- 0.2 in Hi-PN, 6.1 +/- 0.2 in I-PN, and 5.6 +/- 0.1 mg.kg-1.min-1 in Lo-PN. Insulin sensitivity was assessed with the euglycemic clamp using two insulin infusion rates: 2 and 4 mU.kg-1.min-1. The rate of glucose disappearance was 14.8 +/- 0.4 and 25.3 +/- 0.7 in Hi-PN, 15.3 +/- 0.4 and 26.9 +/- 0.5 in I-PN, and 16.1 +/- 0.6 and 31.5 +/- 0.5 mg.kg-1.min-1 in Lo-PN, respectively. HGP was suppressed by 86 in the 2- and by 90% in the 4-mU insulin clamp in the Lo-PN, whereas HGP was suppressed by 45 and by 79% in the two steps in the Hi-PN group.(ABSTRACT TRUNCATED AT 250 WORDS)