PurposeTo test spatial ability in athletes with different axial rotation experience and analyze their behavioral data to explain the cognitive mechanisms of spatial ability in athletes.MethodsExperiment 1: A total of 147 athletes were selected for the paper-and-pencil mental rotation test (MRT). The athletes were separated according to three sport types: open high-spatial (OH) sport, closed high-spatial (CH) sport, closed low-spatial (CL) sport. Spatial ability testing with a two-factor mixed experimental design of 3 (sport type) × 2 (stimulus type). Experiment 2: In this study, 47 players were selected for computerized mental rotation test, with a three-factor mixed experimental design of 3 (sport type) × 2 (angle: 45°, 90°) × 3 (rotational axis: left–right axis, up–down axis, and front–back axis). Repeated-measures ANOVA was performed to evaluate the data.Results(1) The CH group and OH group outperformed the CL group in the non-embodied task (all ps < 0.003) and the CH group was better than the other groups in the embodied and tasks (all ps < 0.008). (2) Under 45° rotational conditions, the reaction time (RT) for the left–right (LR) and up–down (UD) axes were shorter than that for the front-back (FB) axis (all ps < 0.026). However, under 90° conditions, the RT for FB < LR < UD, with superior accuracy and rotational speed for the FB axis than for the LR and UD axes (all ps < 0.034). (3) Male players from the CH and CL groups had shorter RTs than did those from the OH group at both angles (all ps < 0.047). For female players, the CH group presented a shorter RT than the OH and CL groups did at 90° (all ps < 0.006). (4) No sex difference was found for paper and pencil MRTs, but a male advantage existed only in the CL group for computerized MRTs (p = 0.005).ConclusionThe motor skills associated with axial rotation could promote mental rotation performance and compensate for sex differences in mental rotation ability.