Although conclusive evidence has been obtained for the presence of radiation-induced bystander effects, the mechanisms that trigger and regulate these processes are still largely unknown. The bystander effect may play a critical role in determining the biological effectiveness of low-dose exposures, but questions on how to incorporate it into current models and extrapolate the risks of radiation-induced carcinogenesis are still open. The Gray Cancer Institute soft X-ray microbeam has been used to investigate the dose-response relationship of the bystander effect below 0.5 Gy. The survival response of V79 cells was assessed after the irradiation of a single cell within a population with a submicrometer-size beam of carbon K X rays (278 eV). Above 0.3 Gy, the measured bystander cell killing was in agreement with previously published data; however, a significant increase in the scatter of the data was observed in the low-dose region (<0.3 Gy). The data distribution observed indicates a binary behavior for triggering of the bystander response. According to our hypothesis, the probability of triggering a bystander response increases approximately linearly with the dose delivered to the single selected cell, reaching 100% above about 0.3 Gy. The magnitude of the bystander effect, when triggered, is approximately constant with the dose and results in an overall approximately 10% reduction in survival in our system. This suggests that the event that triggers the emission of the bystander signal by the hit cell is an all-or-nothing process. Extrapolation of the data indicates that when a single fast electron traverses a V79 cell, there is a probability of approximately 0.3% that the cell will emit the bystander signal. The data presented in this paper have also been analyzed statistically to test the possibility that complex DNA double-strand breaks may be the initial critical event.