Hyperinsulinemia is associated with an adverse pattern of cardiovascular risk factors, including obesity, elevated triglyceride levels, low levels of high-density lipoprotein (HDL) cholesterol, and elevated blood pressure. Whether hyperinsulinemia precedes (and perhaps causes) this deterioration in the risk factors or merely accompanies the deterioration is controversial. We therefore examined the 8-year changes in lipids, lipoproteins, and blood pressure as a function of baseline levels of fasting insulin in 1,383 nondiabetic Mexican-American and non-Hispanic white subjects enrolled between October 1979 and November 1982 in the San Antonio Heart Study, a population-based longitudinal study of cardiovascular risk factors and diabetes in San Antonio, Texas. After age and concomitant changes in body mass index were adjusted for, fasting insulin at baseline was found to be correlated positively with 8-year changes in triglyceride levels and negatively with 8-year changes in HDL cholesterol levels (p less than 0.05). Among the non-Hispanic whites, insulin was more strongly correlated with a decline in HDL cholesterol levels in women than in men (p less than 0.001). Fasting insulin was also positively correlated with changes in both systolic and diastolic blood pressure in non-Hispanic whites, but not in Mexican Americans, although these correlations were slightly diminished and no longer achieved statistical significance after subjects receiving antihypertensive medications were excluded. These results support the hypothesis that in nondiabetic subjects, insulin has a direct regulatory effect on triglyceride and HDL cholesterol levels. These data provide evidence for a possible role for insulin in blood pressure regulation, at least in non-Hispanic whites, although further analysis of this issue is warranted.