This study examined the relation between control of motor imagery and generation and transformation of visual imagery by testing 54 subjects. We used two measures of the Controllability of Motor Imagery test to evaluate the ability to control motor imagery. One was a recognition test on which the subject imagines as if one sees another's movement, and the other was a regeneration test on which one imagines as if one moves one's own body. The former test score was related to processing time of a mental rotation task and the latter one was not but would reflect sport experience. It was concluded that two meanings of the test could reflect different aspects such as observational motor imagery and body-centered motor imagery.