A physiological compartmental model of alpha-linolenic acid metabolism was derived from the plasma concentration-time curves for d5-18:3n-3, d5-20:5n-3, d5-22:5n-3, and d5-22:6n-3 in eight healthy subjects. Subjects received a 1-g oral dose of an isotope tracer of alpha-linolenate (d5-18:3n-3 ethyl ester) while subsisting on a rigorously controlled beef-based diet. By utilizing the Windows Simulation and Analysis Modeling program, kinetic parameters were determined for each subject. Half-lives and mean transit times of the n-3 fatty acids in the plasma were also determined. The model predicted plasma values for the n-3 fatty acids in good accordance with the measured steady state concentrations and also predicted dietary linolenic acid intake for each subject in accordance with values determined by lipid analysis of the diet. Only about 0.2% of the plasma 18:3n-3 was destined for synthesis of 20:5n-3, approximately 63% of the plasma 20:5n-3 was accessible for production of 22:5n-3, and 37% of 22:5n-3 was available for synthesis of 22:6n-3. The inefficiency of the conversion of 18:3n-3 to 20:5n-3 indicates that the biosynthesis of long-chain n-3 PUFA from alpha-linolenic acid is limited in healthy individuals. In contrast, the much greater rate of transfer of mass from the plasma 20:5n-3 compartment to 22:5n-3 suggests that dietary eicosapentaenoic acid may be well utilized in the biosynthesis of 22:6n-3 in humans.