While testing for genotoxicity is usually performed on single chemicals, exposure of humans often involves combinations of agents. Previous results from this laboratory showed supra-additivity for the induction of micronuclei in p53-mutated mouse lymphoma L5178Y cells after combined treatment with gamma-radiation from a 137Cs source and ethyl methanesulfonate (EMS). The question now was whether supra-additivity was a general phenomenon for the genotoxicity of this combination of a physical and a chemical DNA-damaging agent or whether the result was species- and cell type-specific. The same combination of agents was investigated in two human lymphoblastoid cell lines, TK6 (wild-type p53) and WTK1 (mutated p53), and primary fibroblasts from a fetal human lung. Doses were in the linear dose-effect range, resulting in a 1.5- to 3-fold increase in micronuclei above control. Radiation doses were between 125 and 350 mGy, while the EMS concentrations were 20-50 microg/ml for the cell lines and 250-350 microg/ml for the primary cells. In none of the human test systems was supra-additivity observed. With the WTK1 cells, which are most similar to the mouse cells regarding p53 status, there was even a tendency for a sub-additive combination effect. Possible explanations for the difference to the mouse cells could be related to species-specific aspects, different consequences of the p53 mutations or the presence of additional mutations. It is concluded that caution is advised in the interpretation and extrapolation of experimental results of mixture toxicity data because the outcome could be highly specific for the given selection of agents, doses and assays.