The relationship between blood ethanol concentration and hepatic ethanol metabolism commonly is calculated using the Michaelis-Menten equation and a one-compartment model that assumes equality of blood and hepatic ethanol concentrations. However, at low blood concentrations, most of the ethanol arriving at the liver is metabolized, and hepatic ethanol concentrations may fall far below that of the entering blood. We have developed a two-compartment model of ethanol metabolism that accounts for the fall in ethanol concentration that may occur as blood traverses the liver and used this model to make predictions concerning ethanol metabolism at various blood ethanol concentrations. The two-compartment model predicts that near-complete saturation will occur more abruptly and at a lower blood concentration (∼3 mM) than is the case with the one-compartment model. Thus, the two-compartment model predicts a near-constant ethanol elimination rate for blood ethanol concentrations above 3 mM (as commonly observed in human subjects), whereas the one-compartment model predicts an increasing elimination rate over the range of concentrations observed in experimental studies. In agreement with observed data, the two-compartment model predicts that first-pass metabolism should be extremely sensitive to the rate of ethanol absorption. Application of this model to previously published data indicated that, when absorption was slowed via concomitant food ingestion, first-pass metabolism accounts for ∼50% and 10% of ethanol dosages of 0.15 g/kg and 0.3 g/kg, respectively. When ingested without food, there is negligible first-pass metabolism of even very small ethanol dosages (0.15 g/kg). These findings suggest that first-pass metabolism is an unimportant determinant of the blood ethanol response to ingestion of potentially inebriating doses of ethanol.