1 Duplex scanning was used to measure liver blood flow (hepatic artery and main branches of the portal and hepatic veins) in six healthy subjects, five cirrhotic patients, and six hepatitis patients. Antipyrine clearance and formation clearances to its metabolites were also measured. 2 Compared with healthy control subjects, cirrhotic patients had a lower hepatic vein blood flow (-76%, P < 0.05). This was due primarily to a lower portal vein blood flow (-36%, NS). A statistically significant difference in liver blood flow between patients with hepatitis and normal subjects was not detected. 3 Antipyrine half-life, clearance, and the area under the serum drug concentration vs time curve were significantly different in cirrhotic patients compared with the healthy subjects (mean ± s.d.-healthy controls: t, 2 = 13.7 ± 3.0 h, CL = 30.0 ± 8.6ml h-' kg-', AUC = 549 ± 139 mg 1-1 h; cirrhotic patients: ti,2 = 32.4 ± 1.7 h, CL = 12.3 ± 2.1 ml h-1 kg-', AUC = 1061 ± 218 mg 1-1 h; P < 0.008). Antipyrine halflife, clearance, and the area under the serum drug concentration vs time curve were not significantly different in hepatitis patients compared with the healthy subjects (hepatitis patients: tb = 14.3 ± 3.7 h, CL = 29.3 ± 8.5 ml h-1 kg-I, AUC = 498 ± 142 mg I1-h). The volume of distribution of antipyrine was similar in all three groups of subjects. 4 Formation clearances to 3-OH-methylantipyrine, 4-OH-antipyrine, and norantipyrine were all lower (P < 0.009) in cirrhotic patients compared with normal subjects and hepatitis patients. However, only the formation clearance to norantipyrine was lower (P < 0.002) in hepatitis patients when compared with normals. 5 There were strong correlations between the logarithmic transformations of the formation clearances for all antipyrine metabolites (r = 0.88-0.93, P < 0.0001). However, the only orthogonal regression line with a slope not significantly different from one was that for norantipyrine and 3-OH-methylantipyrine indicating that these two metabolites may be metabolized by a common enzyme(s). 6 Weak correlations were found between the logarithmic transformations of hepatic vein blood flow and antipyrine clearance and the logarithmic transformations of hepatic vein blood flow and formation clearances to antipyrine metabolites (r= 0.54-0.62, P < 0.05). 7 Cirrhotic patients may have difficulty metabolizing drugs with high hepatic extraction ratios or drugs that are eliminated by the same liver enzyme systems which produce the three major metabolites of antipyrine. Hepatitis patients may have difficulty metabolizing drugs that share the same hepatic enzyme system responsible for the formation of norantipyine from antipyrine.