To determine the preferred infusion-sampling mode for isotopic studies of free fatty acid (FFA) metabolism, tracer [( 14C]palmitate) was infused into the left ventricle of five anesthetized dogs, and tracee ([3H]palmitate) was infused into three separate peripheral veins of each dog. The [14C]palmitate specific activity (SA) was lower in mixed venous than arterial blood, and [3H]palmitate SA was equal in both sites. The actual infusion rate of [3H]palmitate [2.15 +/- 0.31 X 10(5) disintegrations/min (dpm).kg-1.min-1] could be accurately predicted (2.14 +/- 0.32 X 10(5) dpm.kg-1.min-1) using the known [14C]palmitate infusion rate and the arterial plasma [14C]-to-[3H]palmitate ratio. In contrast, the mixed venous [14C]-to-[3H]palmitate ratio resulted in overestimates (P less than 0.05) of the actual [3H]palmitate infusion rate. In summary, venous tracer infusion with arterial blood sampling for FFA tracer studies provides the most accurate estimates of tracee rate of appearance.