Leucine influx into six brain regions was determined in anesthetized rats with the in situ brain perfusion technique using either saline or plasma perfusate. This technique has several advantages over other methods such as the brain uptake index (BUI) technique. The concentration dependence of L-leucine influx was best described by a model with a saturable and a nonsaturable component. For the saturable component, Vmax equaled 10.67 +/- 0.21 X 10(-4) mumol s-1 g-1 and Km equaled 0.0257 +/- 0.0009 mumol ml-1, whereas the constant of nonsaturable diffusion (Kd) equaled 0.957 +/- 0.067 X 10(-4) s-1 in the parietal lobe during saline perfusion. Vmax was higher in the cortical lobes than in other brain areas, probably owing to a greater capillary surface area. There were no regional differences in Km or Kd. The apparent Km for L-leucine during plasma perfusion was 20 times greater than the Km during saline perfusion, and 3 to 4 times greater than the plasma leucine concentration, owing to competitive inhibition of leucine transport by other large neutral amino acids in plasma. These results for Vmax, Km, and Kd differ by three- to fourfold from previous estimates obtained with the BUI technique. The high apparent Km during plasma perfusion indicates that leucine influx is a linear function of plasma concentration up to 0.5 mumol ml-1 when the plasma concentrations of other amino acids remain constant, whereas influx would be approximately constant when plasma concentrations of all large neutral amino acids increased or decreased by a constant fraction.