Inert gas clearance has been used for 20 years to measure hepatic blood flow. Injection of a saline solution of 85Kr or 133Xe is usually made via the PV, and the resulting hepatic clearance is monitored with a Geiger-Müller tube, scintillation crystal, or gamma camera. Complex slow components in 133Xe clearance curves, once believed to indicate a correspondingly complex hepatic microcirculation, are now considered to be caused by nonhepatic radioactivity. Normal liver is therefore believed to receive a homogeneous perfusion throughout the depth of tissue in any given region. HA blood and PV blood are normally completely mixed in the hepatic sinusoids. Macroscopic variations in tissue perfusion may exist in different lobes of the liver in both animals and man. The technique expresses flow in units of milliliters per minute per 100 g. Accurate and acceptably reproducible results have been obtained after PV injection of isotope; fast component analysis of 133Xe clearance is most appropriate, while beta detection of 85Kr yields a simple monoexponential curve. Normal hepatic blood flow in dogs and in man is 100-130 ml min-1 100 g-1. Employing sites of isotope administration other than the PV produces inaccurate results unless appropriate corrections are made. Accuracy of flow measurement is critically dependent on a knowledge of the partition coefficient of the gas used. Liver disease per se does not affect measurement accuracy, and many practical features make the technique an attractive tool for the measurement of hepatic hemodynamics in man. Nevertheless, it is essential that the investigator be aware of certain limitations of the method, and carefully apply current concepts of clearance curve analysis and interpretation, in order to derive maximum advantage.