Fifty out-patients with endogenous hypertriglyceridemia were submitted to glycosylated hemoglobin (GHb) and glucose tolerance assessment. Fifteen had normal glucose tolerance (NGT), 15 had impaired glucose tolerance (IGT) and 20 had non-insulin-dependent diabetes mellitus (NIDDM). GHb was 6.3% in NGT, 7.3% in IGT and 8.11% in NIDDM and was significantly correlated to fasting and post-prandial plasma glucose (p less than 0.001) in NIDDM group and to peak, area (p less than 0.001) and 2-h plasma glucose levels (p less than 0.05) of OGTT in the NGT and IGT groups. Out of the 15 IGT subjects only 6 had GHb levels above the control range, while 9 had normal GHb values. These data show that also in hyperlipemic subjects GHb values are related to glucose tolerance, and suggest that GHb evaluation alone is not sufficient for the diagnosis of impaired glucose tolerance. In order to evaluate whether plasma turbidity can affect GHb dosage, GHb was evaluated in 10 hyperlipemic subjects with various degrees of hypertriglyceridemia both in the presence (whole blood hemolysate) and absence (isotonic saline hemolysate) of plasma triglycerides. The results show that, with our method, plasma turbidity does not affect GHb evaluation.