We tested the hypothesis that the previously observed association among hypertriglyceridemia, hyperinsulinemia, and insulin resistance could be explained by a defect in insulin's antilipolytic effect. Insulin action was measured in 10 nondiabetic and 8 diabetic patients with hypertriglyceridemia (fasting plasma triglyceride 800 ± 154 and 1105 ± 445 mg/dl, respectively, P NS; fasting plasma glucose 99 ± 3 and 161 ± 12 mg/dl, respectively, P < .001) and in 8 weight-matched normolipemic nondiabetic individuals (fasting plasma triglyceride and glucose 110 ± 21 and 91 ± 3 mg/dl). The slope of the decay in plasma free fatty acid (FFA) during insulin infusion was used as an index of insulin's antilipolytic effect. Insulin stimulation of glucose uptake in vivo during intravenous hyperinsulinemic clamp and in vitro in adipocytes were measures of insulin's glucoregulatory action. Both glucoregulatory and antilipolytic effects were similarly reduced in both hypertriglyceridemic groups compared with normal subjects. The plasma triglyceride concentration correlated positively with the slope of FFA suppression by insulin (r = .81, P < .0001) and the fasting FFA concentration (r = .65, P < .0001). In multiple linear regression analysis, insulin's antilipolytic effect and the fasting FFA concentration explained 83% of the variation in the plasma triglyceride concentration. These associations were independent of insulin's glucoregulatory effect and the fasting plasma insulin concentration. The data indicate that patients with endogenous hypertriglyceridemia are resistant to both the antilipolytic and glucoregulatory actions of insulin and that increased flux of FFA as a result of the latter, rather than hyperinsulinemia, is responsible for elevation of very-low-density lipoprotein production. Finally, we suggest that previous inconsistent results regarding insulin's antilipolytic effect in obesity and non-insulin-dependent diabetes mellitus could be related to variation in the triglyceride concentrations of the study populations.