Survival of HT29 cells was measured after irradiation with single doses of X-rays (0.05-5 Gy) and neutrons (0.025-1.5 Gy), using a Dynamic Microscopic Imaging Processing Scanner (DMIPS) with which individual cells can be accurately located in tissue culture flasks, their positions recorded, and after an appropriate incubation time the recorded positions revisited to allow the scoring of survivors. The response over the X-ray dose range 2-5 Gy showed a good fit to a Linear-Quadratic (LQ) model. For X-ray doses below 1 Gy, an increased X-ray effectiveness was observed with cell survival below the high-dose LQ prediction. The value of --dose/loge (SF) for each experimental data point, plotted against dose, demonstrated clearly how X-rays are maximally effective at doses approaching zero, becoming less effective as the dose increases and with minimal effectiveness at about 0.6 Gy then becoming more effective again as the dose increases above 1.5 Gy. This phenomenon was not seen with neutrons. Neutron RBE was calculated for each X-ray data point by taking each X-ray survival value and comparing it with the common LQ fit to all the neutron data. Over the X-ray dose range 0.05-0.2 Gy, the RBE is close to 1 indicating that these very low doses of X-rays are of similar effectiveness to neutrons in killing cells. The increase in RBE with increasing dose over the range 0.05-1 Gy, and the slight decrease in RBE above 1 Gy, reflect primarily the changes in X-ray sensitivity over the whole dose range of 0.05-5 Gy. Several arguments suggest that this phenomenon could reflect an induced radioresistance so that in this system low single doses of X-rays are more effective per Gy than higher doses in reducing cell survival because only at higher doses, above a threshold, is there sufficient damage to trigger radioprotective mechanisms.