Little is known about the performance of the laser Doppler flowmeter (LDF) with changes in flow characteristics in the hepatic microcirculation. Red blood cell (RBC) flux, as measured by LDF, is sensitive to alterations in RBC velocity and RBC concentration, and both parameters must be considered when evaluating the technique. In vitro, linearity of LDF signal with RBC velocity up to 4 mm/s (r greater than 0.99) and RBC volume fractions up to 1.5% were demonstrated (r greater than 0.98). At higher velocities and volume fractions, RBC flux was underestimated. In portally perfused rat liver, LDF output was linearly related to total liver blood flow (TLBF; r greater than 0.9) for perfusate hematocrits between 2.5 and 40%, although the slope varied between preparations. At constant TLBF (2 ml.min-1.g-1), the LDF output changed linearly with perfusate hematocrit up to 20% (r greater than 0.99) but underestimated the increase at 40%. These results suggest that the LDF responds linearly to velocity-mediated flow changes but that it may underestimate a change if mediated through alteration in tissue RBC concentration. With hepatic nerve stimulation, TLBF fell to 51 +/- 14% of prestimulation (P less than 0.001), whereas the LDF output and superficial flow measured by 85Kr clearance fell to 12 +/- 10 and 14 +/- 10% (both P less than 0.01 vs. TLBF), respectively, suggesting that the LDF may be used to follow rapid flow changes in the periphery of the liver.