Isotopic glucose turnover was measured during euglycemic glucose clamp studies in spontaneously hypertensive rats and normotensive Wistar-Kyoto control rats, under pentobarbital sodium anesthesia. At an insulin infusion rate of 4 mU · kg⁻¹ · min⁻¹, glucose disposal rate and hepatic glucose output were not significantly different in either group, at a plasma glucose of 6 mM. However, steady-state plasma insulin levels in spontaneously hypertensive rats were more than double those in Wistar-Kyoto rats (2.34 ± 0.16 [mean ± SE] vs. 6.78 ± 0.58 nM, P < 0.00005, n = 7 in each group). Additional studies (n = 30 in spontaneously hypertensive rats and n = 32 in Wistar-Kyoto control rats) were conducted to match insulin levels in the two groups over a wide range, using infusion rates of 0.5–10 mU · kg⁻¹ · min⁻¹. When the responses of glucose disposal rate and hepatic glucose output to insulin were analyzed from the standpoint of plasma insulin levels, a highly significant difference was seen in the response to insulin between spontaneously hypertensive rats and Wistar-Kyoto rats, for both glucose disposal rate (P < 0.000005) and hepatic glucose output (P = 0.00007). The slope of the regression line for glucose disposal rate versus plasma insulin was lower in spontaneously hypertensive rats compared with Wistar-Kyoto rats (2.652 vs. 4.864, P < 0.001), indicating that insulin stimulation of glucose disposal rate was reduced by 50% in spontaneously hypertensive rats. The slope of hepatic glucose output versus plasma insulin was 0.739 in spontaneously hypertensive rats, indicating impaired suppression of hepatic glucose output, compared with Wistar-Kyoto rats, where the slope was –10.3, (P < 0.001 vs. spontaneously hypertensive rats). The clearance of insulin showed great variability in both groups but was much lower in spontaneously hypertensive rats compared with Wistar-Kyoto rats (6.7 ± 0.9 vs. 14.9 ± 1.3 ml · kg⁻¹ · min⁻¹, P < 0.000005). Post hoc power analysis showed that the within-group variation resulted in low power and high risk of type II errors when comparisons were restricted to a small sample size (n = 8). We conclude that insulin sensitivity is impaired in spontaneously hypertensive rats when expressed as a function of the ambient plasma insulin level, but not when expressed as a function of the administered dose of insulin, because of the risk of type II errors caused by the great variability in insulin concentrations at any given insulin infusion rate when the sample size is small.