Morphological and functional heterogeneity of hepatocytes according to their position in the liver lobule has been known for many years. The digitonin-collagenase perfusion technique is widely used to study hepatocyte heterogeneity and has yielded reliable data. However, with this procedure, periportal (PP) or perivenous (PV) hepatocytes are isolated from different livers, allowing only comparison between cell populations issued from two separate animals. To overcome this drawback, we have modified this technique by perfusing the two main rat liver lobes of a single animal in succession. The procedure involved alternate clamping of the median and the left lateral lobes, restricting digitonin infusion to one lobe via the portal vein, and to the other via the caudal vena cava. Lobe exclusion during digitonin perfusion, and zonal restriction of digitonin-induced damage, were monitored using macroscopic and histological controls. We compared our results with previous data on PP and PV hepatocytes issued from two different livers using the conventional digitonin-collagenase perfusion technique. First, we found that the cellular sensitivity to angiotensin II, a calcium-mobilizing agonist, was 60% to 80% higher in PV than in PP hepatocytes, whereas, previously, no difference had been recorded. Second, we found that albumin messenger RNAs (mRNAs) were 35% more abundant in PP than in PV hepatocytes, whereas, previously, larger differences had been reported. Our results show that PP and PV hepatocytes may be isolated from a single liver using an improved digitonin-collagenase perfusion technique. Furthermore, we suggest that zonal differences can be artificially masked or amplified when comparing PP and PV cell populations from two different livers, indicating that it is preferable to use a single liver for accurate zonal comparisons between hepatocytes.