Many strains of mice have relatively high background incidences of liver cell tumours which can be readily increased under various experimental conditions. In contrast, in rats the background incidence of liver tumours is negligible or very low. In the present study the role of various environmental factors in liver tumour development in mice was investigated. The susceptibility of liver cells, in a state of replication, to undergo neoplastic change following exposure to certain chemicals was investigated in mouse and rat. Liver cell ultrastructure and histochemical distribution of enzymes were also studied in an attempt to characterise mouse liver tumours. Since exposure of mice to various chemicals, that cause induction of the microsomal enzymes, can result in increased incidences of liver cell tumours, the possible role of naturally occurring inducers in the laboratory rodent's environment, was investigated. Liver microsomal enzyme induction followed exposure to conventional diet and sawdust bedding, but was less pronounced than that caused by dieldrin, which was included as a positive control. Unlike dieldrin, microsomal enzyme inducers present in conventional diet and sawdust bedding did not increase the background liver tumour incidences. Liver tumour development, in some mice housed on filter-paper and fed semi-synthetic diet, indicated the presence of a pre-existing oncogenic factor. Ultrastructural and histochemical observations revealed no specific characteristics that distinguished dieldrin exposed mouse liver tumour cells from non-tumour cells, control group mouse liver tumour cells or non-tumour cells. Dieldrin exposed mice surviving beyond 18 months showed higher incidences of hepatocellular carcinomata and lung metastases compared with mice killed before 18 months. High incidences of liver cell tumours developed in aflatoxin B[1]-exposed, partially-hepatectomised mice in contrast to non-partially hepatectomised control or aflatoxin B[1]-exposed mice. Dieldrin exposure of partially hepatectomised mice was associated with earlier death due to liver tumours when compared with mice exposed to dieldrin but not subjected to partial hepatectomy. In contrast exposure of partially hepatectomised rats to dieldrin did not result in liver tumour development. It is concluded that partial hepatectomy followed by exposure to a microsomal enzyme inducer promotes the development of liver tumours in mice having pre-existing but unknown oncogenic factors involving the liver. It has also been shown that partial hepatectomy has a promoting action in mice exposed to an initiating carcinogenic chemical but no significant effect on the incidences of liver tumours in control group mice.