A major advance in the study of mature hepatocytes ex vivo was the development of methods for dispersing the liver with proteolytic enzymes and harvesting intact viable cells. Howard and coworkers introduced bacterial collagenase for this purpose,'an approach that was superceded in the intact organ by collagenase perfusion'because of substantially increased yields of viable cells by this method. This was followed by the demonstration that collagenase-isolated he atoc tes, in the appropriate nutrient medium, could be established in c~ lture.~'Because the perfusion technique made available large numbers of cells, preparation of mass primary cultures was possible, an approach that has attracted attention because it involves a homogeneous preparation of hepatic parenchymal cells and also sustains cell viability well beyond that obtained with other methods for mature differentiated hepatocytes.'Primary hepatocyte cultures can be prepared with nonproliferating cells (from normal liver) and are established within 4 h of cell plating (see below). This proximity, both physiologically and temporally, to the intact liver has encouraged numerous studies of the retention of liver-specific function in culture-a problem raised frequently in the past with respect to proliferating cell lines from liver. 5.'The first studies of adult rat hepatocytes in primary monolayer culture were concerned with quantitation of specific function, comparing cultured hepatocytes to the intact parent liver, and demonstrated diminished specific function in some instance^.^'Subsequent studies in several laboratories have established that phenotypic change occurs in primary hepatocyte culture, not only in mammalian but also in amphibian hepatocytes7; it is distinct from a process of senescence or impaired viability; and it may be quantitatively impressive, involving key hepatic functions. The discussion that follows will be concerned with the documentation of phenotypic change in various hepatocyte culture systems, with the available information concerning its prevention or reversal and with its implications for studies using hepatocyte culture. The scope of this review will be limited largely to primary culture of mature hepatocytes, to the exclusion of fetal hepatocyte culture and propagated