Deer behavioural responses to culling, such as transient home range shifts and changes in diel activity patterns, which are factors reducing capture efficiency, become obstacles to achieving long-term management sustainability. However, most studies have analysed these responses on a long time-scale. Furthermore, the differences among various culling methods (e.g., driving versus stalking) have not been evaluated. In this study, photographic data from 28 camera traps of a sika deer (Cervus nippon) population in the Tanzawa Mountains, Japan, were analysed to evaluate differences in behavioural responses to different culling methods. First, to compare the duration of transient home range shift between two culling methods, a Cox proportional hazard model was applied to the day period encompassing when culling began to when a deer was re-photographed at a culling site; we found driving required slightly longer days. Second, temporal changes in deer diel activity patterns between disturbed (seven days after culling) and undisturbed (out of the disturbed period) periods were evaluated with kernel density estimation for each culling method. The degree of change in diel activity pattern between the two periods was almost the same between culling methods; however, peak activity during the disturbed period differed between the two culling methods, probably owing to the difference in start time. We conclude that deer immediately changed their behaviour according to the culling method characteristics; therefore, deer behavioural change over a short time period should be incorporated into the planning of population control.