Insulin resistance is associated with a mild hyperviscosity syndrome, which is more closely related to insulin resistance than to the clinical scoring of the metabolic syndrome. In studies using the intravenous glucose tolerance test with minimal model analysis we reported that low insulin sensitivity (SI) is associated with increased erythrocyte aggregability (EA). Actually, this issue is confusing because insulin resistance is often associated with compensatory hyperinsulinemia (insulin being a hormone with reported hemorheologic effects) and that a decline in insulin secretion has marked metabolic effects that modify blood rheology. From the intravenous glucose tolerance test (IVGTT) the minimal model allows the calculation of SI, insulin response, and an overall glucose tolerance parameter termed "disposition index" (DI) that measures whether insulin response is adequate or not for the level of insulin sensitivity. In this study we assessed SI, insulin response, and DI during an IVGTT in 335 subjects of both genders (age 8-77 yr; BMI 14-67 kg/m2). SI was only correlated (negatively) with EA (Myrenne M r = -0.285; p = 0.0001; M1 r = -0.240 p = 0.003). Fasting insulin was also correlated (positively) with EA (Myrenne M r = 0.233, 0.00880; M1 r = 0.320 p = 0.0003; SEFAM TA r = -0.342 p = 0.04; SEFAM S60 r = 0.419 p = 0.01) and SEFAM RBC disaggregation thresholds (γS = r = 0.372 p = 0.025; γD = r = 0.504 p = 0.002). Fasting DI (SI × fasting insulin) is negatively correlated to M (r = -0.274; p = 0.002) and M1 (r = -0.225; p = 0.01) but also positively to whole blood viscosity (r = 0.168; p = 0.01) and hematocrit (r = 0.142; p = 0.05). Stimulatory DI (SI × insulin peak) fails to be correlated with any parameter of EA but is negatively correlated to whole blood viscosity (r = -0.150; p = 0.02) and plasma viscosity (r = -0.163; p = 0.01). This study confirms that red cell aggregability is associated with insulin resistance and hyperinsulinemia, but plasma viscosity seems to be more related to overall glucose tolerance than to either SI or insulinemia.