The efficacy of exponentially decreasing low-dose-rate irradiation was compared with that of equivalent doses of multiply fractionated high-dose-rate external-beam irradiation in mice with three different kinds of tumors that varied in terms of alpha/beta ratio, tumor volume doubling times, and cell cycle times. Cell cycle distribution was measured after low-dose-rate and high-dose-rate irradiation. The relative efficacy of low-dose-rate irradiation versus equivalent doses of high-dose-rate irradiation was correlated with the extent of arrest of cells in G2 phase of the cell cycle. Unlike 38C13 murine B-cell lymphoma, neither the HT29 human colorectal xenograft or the SNB75 human glioblastoma xenograft was significantly more sensitive to low-dose-rate than fractionated high-dose-rate irradiation. The 38C13 B-cell lymphoma also had the shortest tumor volume doubling and cell cycle times, the most G2 arrest after low-dose-rate irradiation, more G2 arrest with low-dose-rate than high-dose-rate irradiation, and the highest alpha/beta ratio. The data presented here support the hypothesis that dose-rate effects may be minimal for tumors with small shoulders and large alpha/beta ratios and that arrest of cells in G2 phase plays an important role in cell death mediated by low-dose-rate irradiation. The proliferative rates of tumor cells may modify dose-rate effects further, and the interaction of all of these factors may explain in part the increased efficacy of low-dose-rate irradiation and radioimmunotherapy compared with high-dose-rate irradiation that has been reported in some animal models.