To examine the consequences of liver blood flow variations on drug disposition in cirrhosis, we studied the effects of portacaval shunt on drug clearance in 35 cirrhotic patients. Lidocaine clearance and bioavailability, indocyanine green (ICG) clearance, aminopyrine breath test, and hepatic blood flow were measured before and 18 months after surgery. The patients were divided into two groups according to severity of disease: 14 patients (group 1) had slight liver dysfunction (ICG extraction ratio greater than 0.25) and 21 patients (group 2) had severe liver disease (ICG extraction ratio less than 0.25). After portacaval shunt the decrease in hepatic blood flow was similar for both groups (-65%). In group 1, ICG systemic clearance decreased from 9.10 +/- 0.68 to 4.40 +/- 0.34 ml/min . kg (p less than 0.05), whereas ICG intrinsic clearance remained unchanged; lidocaine systemic clearance decreased from 7.93 +/- 0.93 to 5.09 +/- 0.33 ml/min . kg (p less than 0.05), whereas lidocaine intrinsic clearance remained unchanged; bioavailability increased from 0.601 +/- 0.076 to 1, resulting in an abrupt reduction of oral clearance from 18.01 +/- 4.90 to 5.09 +/- 0.33 ml/min . kg (p less than 0.05). In group 2, ICG systemic clearance decreased slightly from 3.90 +/- 0.39 to 2.28 +/- 0.16 ml/min . kg (p less than 0.01) and ICG intrinsic clearance was not modified; lidocaine systemic and intrinsic clearance remained unchanged; and bioavailability increased from 0.779 +/- 0.229 to 1, resulting in a decrease of oral clearance from 7.68 +/- 1.65 to 4.23 +/- 0.37 ml/min X kg (p less than 0.05). The aminopyrine breath test was not affected by surgery in either group. We conclude that reduction of hepatic blood flow after portacaval shunt has only minimal effects on drug disposition in patients with severe liver disease, but results in a notable reduction in the clearance of high-extraction drugs in cirrhotics with mild liver disease.