Regression analyses were performed on individual data from our two previous clinical investigations to establish the cholesterolemic effects of individual fatty acids. Our principal objective was to determine the effects of stearic acid on plasma total cholesterol (TC) and low-density lipoprotein (LDL) cholesterol levels. Our second objective was to determine the variation in the cholesterolemic response of individuals to changes (delta) in the major dietary fatty acids. The best-fitting linear regression equations relating delta TC and delta LDL (mg/dL) were as follows: delta TC = 2.3 delta C14:0 + 3.0 delta C16:0 - 0.8 delta C18:0 - 1.0 delta polyunsaturated fatty acids (PUFA) and delta LDL = 2.6 delta C14:0 + 2.9 delta C16:0 - 0.5 delta C18:0 - 0.7 delta PUFA, where delta fatty acid = change in intake expressed as percent of calories. Based on these equations, in which stearic acid has a significant, negative regression coefficient, and the other regression models analyzed, it appears that stearic acid has an independent cholesterol-lowering effect. Using the equation we developed, 75% of the actual cholesterolemic responses were within +/- 10 mg/dL of the predicted response. In summary, we have developed a predictive equation (similar to those developed by both Keys and Hegsted) to estimate changes in plasma TC and LDL cholesterol levels of young men in response to changes in dietary fatty acids. However, our predictive equation separates stearic acid from the other long-chain saturated fatty acids (SFA) and indicates that it has an independent cholesterol-lowering effect. Thus, stearic acid is a unique long-chain SFA because of its effect on plasma cholesterol level.