Human chylomicron triglyceride (TG) kinetics has been difficult to determine directly owing to technical limitations. This report describes a new method for studying chylomicron metabolism. Healthy volunteers (n = 10) sipped a drink providing 175 mg fat x kg(-1) h(-1) for 7.5 h to produce a steady-state chylomicronemia. A commercial 10% intravenous lipid emulsion was labeled with [3H]triolein, purified by high-performance liquid chromatography, and sterilized. A trace amount of labeled emulsion was injected intravenously 30 min before (i.e., in the fasting state) and 5, 6, and 7 h after sipping began (i.e., triplicate determinations in the fed state). Chylomicron half-lives were calculated from the monoexponential decay curves, and apparent distribution volumes were estimated by back-extrapolation to time zero. Plasma and estimated chylomicron TG concentrations increased from 89+/-13 and 0.8+/-0.3 to 263+/-43 and 91+/-39 mg/dL (mean +/- SEM), respectively, with feeding. Tracer-determined chylomicron TG half-lives were 5.34+/-0.58 and 6.51+/-0.58 min during the fasting and fed states, respectively (P < 0.01). The apparent distribution volume during the fasting state was 24% greater than plasma volume (4515+/-308 vs. 3630+/-78 mL, P < 0.02), consistent with significant margination of lipid emulsion particles to endothelial binding sites. Margination was reduced during the fed state, suggesting that native chylomicrons competed with lipid emulsion particles for endothelial lipoprotein lipase. The results indicate that a radiolabeled commercial lipid emulsion is metabolized in a fashion similar to native chylomicron TG, and thus can be used to study chylomicron TG kinetics.