Besides being excreted into bile, several anionic drugs and anionic drug conjugates are excreted from liver into plasma by a carrier-mediated process (sinusoidal efflux). In the study presented here, we investigated the influence of albumin on the net sinusoidal efflux of an organic anion from isolated perfused liver, with net sinusoidal efflux defined as the resultant of efflux and reuptake. We tested the hypothesis that albumin exerts its effect on net sinusoidal efflux through inhibition of repeat uptake rather than on the sinusoidal efflux process itself. We also studied possible acinar heterogeneity in net sinusoidal efflux rate of the organic anion. Isolated rat livers were preloaded with the nonmetabolizable anionic model compound dibromosulfophthalein in the absence (type I) or presence (type II) of bovine serum albumin. Fluorescence microscopy of dibromosulfophthalein in liver sections showed selective zone 1 loading after type I loading and homogeneous acinar distribution after type II loading. The rate of net sinusoidal efflux and biliary excretion was then studied in single-pass perfusions with various concentrations of bovine serum albumin in the medium. Net sinusoidal efflux of dibromosulfophthalein showed a marked dependence on albumin concentration in both type I and type II experiments. Net efflux rate reached a maximal value at a medium concentration of 300 μmol/L bovine serum albumin and was not increased further at 600 μmol/L. The initial uptake rate of dibromosulfophthalein was not significantly different in the 300 μmol/L and 600 μmol/L bovine serum albumin experiments. We conclude that the net sinusoidal efflux rate of the organic anion from the hepatocyte as measured in perfusion outflow medium is the resultant of an albumin-independent sinusoidal efflux process combined with albumindependent inhibition of reuptake of this material. The relationship between unbound concentration of dibromosulfophthalein in perfusion outflow medium and initial net efflux rate is nonlinear, probably because nonequilibrium conditions exist in the sinusoid, and effluxed dibromosulfophthalein is recaptured by the cells before extracellular binding to albumin can occur. Because no difference was found between efflux in type I and type II perfusions, it was concluded that no acinar heterogeneity exists in transport of dibromosulfophthalein. Surprisingly, the rapid efflux from the cells induced by albumin had virtually no effect on the biliary excretion rate of dibromosulfophthalein. Therefore we conclude that the sinusoidal efflux process is operating from an intracellular pool different from that for biilary excretion.