Based on a previous study with 8 chemicals (Müller et al., 1993) the applicability of the in vitro rat hepatocyte micronucleus assay was evaluated by testing a further 21 compounds of different chemical classes. The obtained results are in good agreement with the known genotoxic profiles of about 90% of the in total tested compounds. Several known mutagens and carcinogens, i.e., alkylating agents, aromatic amines, nitrosamines, nitro compounds, cross-linking agents, and pyrrolizidine alkaloids gave clear positive results in this assay, whereas all of the tested non-carcinogens were negative. The hepatocyte micronucleus assay was shown to distinguish between carcinogenic/non-carcinogenic isomers, such as 2- and 4-acetylaminofluorene (AAF) and 2- and 1-nitropropane (NP). Furthermore, the non-genotoxic nature of several hepatocarcinogens, i.e., the peroxisome proliferating agents fenofibrate, nafenopin, Wy-14,643, diethyl(hexyl)phthalate (DEHP), and the sedative phenobarbital, could be confirmed in this assay. The hepatocarcinogen coumarin exerted mitogenic but no mutagenic properties in the rat hepatocyte micronucleus assay. This compound may act as a liver tumor promoter. Benzo[a]pyrene (B[a]P) and 7,12-dimethylbenzanthacene (DMBA), both belonging to the group of known carcinogenic and mutagenic polycyclic aromatic hydrocarbons, failed to induce micronucleus formation in rat hepatocytes. The high susceptibility of in vitro proliferating hepatocytes to mitotic inhibition, exerted by the strong cytotoxic actions of these compounds, seems to be responsible for these negative results. A strongly reduced mitotic activity can prevent the formation of micronuclei, even when clastogenic effects may have occurred. In the present stage, the in vitro rat hepatocyte micronucleus assay cannot be recommended for screening genotoxicity testing. It should rather be used for special purposes, e.g., when liver-specific mutagenic effects are expected.