We determined whether plasma phospholipid fatty acid levels, an indicator of fatty acid composition in the diet, are associated with fasting serum insulin concentrations (a marker of insulin resistance). We examined cross-sectionally 4,304 middle-aged adults free of diabetes. Plasma fatty acid composition was quantified by gas chromatography. Fasting insulin was strongly and positively associated with the saturated fatty acid percentage in plasma phospholipids, moderately and inversely associated with the monounsaturated percentage, and not appreciably associated with the polyunsaturated percentage. Fasting insulin adjusted for age, smoking status, alcohol consumption, and sports participation, for example, was 29% higher in men and 33% higher in women per 1.9% greater level of saturated fatty acids (the interquartile range). After adjustment for body mass index (BMI) and other covariates, these estimates were 12% and 15% (P < .01 for the difference from zero). A 1.9% greater increment in saturated fatty acid level was also associated with a 2.4-fold higher odds of hyperinsulinemia (fasting insulin > or = 143.5 pmol/L). These data are consistent with studies showing that fatty acid composition of cell membranes modulates insulin action, and support the hypothesis that increased habitual saturated fat intake or a related dietary pattern is a risk factor for hyperinsulinemia.