This study investigated gender differences on the Mental Rotations Test (MRT) as a function of item and response types. Accordingly, 86 male and 109 female undergraduate students completed the MRT without time limits. Responses were coded as reflecting two correct (CC), one correct and one wrong (CW), two wrong (WW), one correct and one blank (CB), one wrong and one blank (WB), or two blank (BB) answers and were analyzed as a function of distractor type (structural or mirror) and occlusion (occluded, non-occluded). The results showed a men advantage on CC, whereas women obtained more BB, CB, and WB outcomes than men did. Gender differences in favor of men were larger for occluded than for non-occluded items on CC. This gender by occlusion interaction was also found for CW and WB. Implications of the results for possible explanation of gender differences on the MRT are discussed. (Contains 3 tables.)