Previous studies have established the reliability of percutaneous portal venous pressure measurement using a Chiba needle, a procedure requiring fluoroscopic guidance. Intrahepatic pressure has been advocated by some as a simple and safe index of portal venous pressure. The aim of this study was to examine the reliability of intrahepatic pressure measurement and its relationship to portal venous pressure. Fifty patients requiring liver biopsy were included: 29 with cirrhosis (n = 20 micronodular, n = 9 macronodular) and 21 with various hepatic disorders but no cirrhosis. The procedure was performed under fluoroscopic guidance, using a Chiba needle connected to a manometer by a saline-filled catheter. Immediately prior to biopsy, each patient underwent measurement of: (i) 3 to 5 separate intrahepatic pressures, the intraparenchymal site being inferred by the lack of blood or bile return; and (ii) portal and hepatic venous pressures, the intravascular position of the needle being ascertained by the reflux of blood and the vessel identified with injection of contrast. Intrahepatic pressure measurements showed great intraindividual variability (variation coefficient up to 115%). Mean intrahepatic pressure (13.19 ± 8.32 mm Hg) was similar to portal venous pressure (14.43 ± 6.10 mm Hg) in the noncirrhotics but significantly lower in the cirrhotics (intrahepatic pressure = 18.34 ± 8.82 mm Hg, portal venous pressure = 22.52 ± 9.47 mm Hg; p < 0.01). The difference between these two parameters exceeded 3 mm Hg in 50% of patients (mean = 9 mm Hg, range = 4 to 19 mm Hg), both in cirrhotics and noncirrhotics. We conclude that intrahepatic pressure measurements are not reproducible and do not represent a reliable index of portal venous pressure.