Induction of micronuclei was used as a measure of the dose-rate effect in Lewis lung carcinoma in vivo. Tumors transplanted on the hind legs of male C57BL mice were irradiated at dose rates of 1 and 0.34 Gy/min, cells were isolated and cultured in vitro, and micronuclei were scored at 24-h intervals. Maximum expression of micronuclei was observed 72 h after plating. The frequency of cells containing micronuclei and the number of micronuclei per single cell were linearly dependent on dose in the range 0-6 Gy. However, a marked inverse dependence on dose rate was observed. The inverse dose-rate effectiveness factor, calculated as the ratio of damage per gray at the lower dose rate to that at the higher dose rate, was 3.25 for frequency of micronuclei and was even higher (4.57) for micronuclei per cell (P < 0.05). Since the differences in exposure time for the different dose rates are not large, our results cannot be explained by the differential effect on cell kinetics during tumor irradiation. However, it cannot be excluded that the differential effect of radiation on division delay and redistribution of cells in the phases of the cell cycle may be expressed during incubation of cells in vitro for micronucleus expression. Furthermore, it can be hypothesized that more cells may die in culture because of interphase death and apoptosis in the higher dose-rate group than in the lower dose-rate group and that these cells were not accessible for the micronucleus assay. The actual explanation for the phenomenon observed requires further experimentation.