Multiple factors may determine insulin resistance and the insulin resistance syndrome. The contributions of genes and environment to the distribution of fasting insulin levels and to the associations of fasting insulin with elements of the syndrome were evaluated in the second examination of the Kaiser Permanente Women Twins Study (Oakland, California, 1989-1990). Subjects included 556 white women (165 monozygous twin pairs, 113 dizygous pairs; 455 women with normal glucose tolerance, 75 with impaired glucose tolerance, and 26 with non-insulin-dependent diabetes by World Health Organization criteria). The intraclass correlation coefficients for log fasting insulin for monozygous and dizygous twin pairs were 0.64 and 0.40, respectively. After adjustment for age, behavioral factors, and body mass index, the estimated classic heritability was 0.53 (p = 0.003). Commingling analysis of fasting insulin indicated the presence of four distributions (p < 0.001), consistent with at least one, and perhaps two, genes influencing this trait. In an unmatched multiple regression model among women from monozygous twin pairs only, log fasting insulin was independently associated with body mass index (p < 0.0001), waist/hip ratio (p = 0.02), and glucose intolerance (p = 0.04), but not with triglycerides, high density lipoprotein cholesterol, or hypertension. After removal of genetic influences by analysis of monozygous intrapair differences, only body mass index (p < 0.0001) remained independently related to fasting insulin. The authors conclude that, in addition to significant genetic influences on fasting insulin, environmental or behavioral factors (particularly nongenetic variation in obesity) are important determinants of fasting insulin and the insulin resistance syndrome.