We used the Chinese hamster ovary cell/hypoxanthine-guanine phosphoribosyl-transferase (CHO/HGPRT) assay to determine the cytotoxicity and mutagenicity of a crude coal oil, the neutral fraction of this crude, and the following three subfractions of the neutral fraction: aliphatic, neutral polar, and a subfraction composed of polycyclic aromatic hydrocarbons plus neutral nitrogen heterocyclics. We also studied the cytotoxicity and mutagenicity of a blend of light and heavy coal-derived fuel oils before and after hydrogenation. All seven mixtures were highly cytotoxic to CHO cells, but the addition of S9 reduced the cytotoxicity. Also, hydrogenation reduced the cytotoxicity of the blend of coal-derived fuel oils. Although highly cytotoxic, none of the seven mixtures induced a clear mutagenic response in the CHO/HGPRT assay. However, previous work has shown that all of the mixtures except the aliphatic subfraction and the blend after hydrogenation are mutagenic in the histidine-reversion assay in Salmonella typhimurium. Based on chemical analyses of the mixtures, the differential sensitivity of Salmonella and CHO cells to nonmutagenic cytotoxins, and studies of the neutral fraction to which additional benzo[a]pyrene had been added, we conclude that the disparity between the results in Salmonella and those obtained in the CHO/HGPRT assay is probably due to the much greater sensitivity of CHO cells (relative to Salmonella) to the cytotoxins in these coal oils. This sensitivity, coupled with the low concentrations of mutagens relative to nonmutagenic cytotoxins in the coal oils, prevents exposure of the cells to concentrations of the mutagens in the mixtures that are high enough to be quantified in the CHO/HGPRT assay.