The reproducibility of hepatic blood flow measurement with pulsed Doppler US was assessed in 6 healthy volunteers. All subjects underwent two different study sessions and were examined by three different operators after overnight fasting. Each operator was blind to his measurements. Hepatic arterial and portal venous blood flow and hepatic arterial maximum velocity were measured; resistive index (RI), pulsatility index (PI) and hepatic perfusion index (HPI) were calculated. Each measurement was repeated three times per session. The data from repeated measurements were analyzed with the analysis of variance (ANOVA) method to assess the intra- and intersubject variations for each variable. The Scheffe test was used to evaluate intra- and interobserver variations. Significant differences among the 6 examined subjects were found for all the variables measured, except for hepatic arterial maximum velocity. The same subject exhibited significant differences in mean velocity of the main portal vein trunk and of its right branch, in maximum hepatic arterial velocity, RI, portal vein section area and blood flow. No significant interobserver variation was observed in the same session and no intraobserver variation detected for all the measurements performed in different sessions. Thus, we conclude that pulsed Doppler US is a repeatable method for measuring hepatic blood flow and may therefore be useful to detect changes in hepatic perfusion.