In cirrhosis, hepatic arterial vasodilatation occurs in response to reduced portal venous blood flow. However, although the hepatic arterial flow reserve is high in patients with cirrhosis, its impact on hepatic function is unknown. This study investigated the effect of adenosine-induced hepatic arterial vasodilatation on different markers of liver function. In 20 patients with cirrhosis (Child-Pugh class A/B/C: n = 2/7/11) adenosine (2-30 microg x min(-1) x kg body wt(-1)) was infused into the hepatic artery and hepatic arterial average peak flow velocities (APV), pulsatility indices (PI), and blood flow volumes (HABF) were measured using digital angiography and intravascular Doppler sonography. Indocyanine green (ICG), lidocaine, and galactose were administered intravenously in doses of 0.5, 1.0, and 500 mg/kg body weight in the presence of adenosine-induced hepatic arterial vasodilatation and, on a separate study day, without adenosine. ICG disappearance, galactose elimination capacity (GEC), and formation of the lidocaine metabolite monoethylglycinxylidide (MEGX) were assessed. Adenosine markedly increased APV and HABF and markedly decreased PI. Serum MEGX concentrations were 63.7 +/- 18.2 (median, 62; range, 36-107) and 99.0 +/- 46.3 (82.5; 49-198) ng/mL in the absence and presence of adenosine infusion, respectively (P =.001). Adenosine-induced changes in MEGX concentrations were correlated inversely to changes in APV (r = -0.5, P =.02) and PI (r = -0.55, P =.01) and were more marked in Child-Pugh class C compared with Child-Pugh class A patients (57.4 +/- 49.9 [44; -14 to 140] vs. 8.4 +/- 16.5 [13; -11 to 35] ng/mL, P <.01). In conclusion, hepatic arterial vasodilatation provides substantial functional benefit in patients with cirrhosis. The effect does not depend directly on hepatic arterial macroperfusion and is observed preferentially in patients with decompensated disease.