The vividness of motor imagery may be assumed to be one of the important variables that may predict the effects of image training on motor skill learning. Likewise, the special characteristics of the vividness of motor imagery are generally assumed to be different from that of the general imagery. In an attempt to construct a test battery to measure the vividness of motor imagery, a respective test (VMI test) was proposed in this study. The first purpose of the present study was to examine the relationships between the vividness of general imagery measured by Betts's test and the vividness of motor imagery by use of factor analytical techniques. The second purpose of this study was to determine the factor structure of the vividness of motor imagery. The third purpose of this study was to test some differences of the factor structure and the factor scores of VMI test between physical education major students (PE) and non-physical education majors (Non-PE). The subjects were 110 male and 80 female undergraduate physical education majors and 107 male and 98 female non-physical education majors. The fifty items of VMI test were constructed under consideration of main sensory modalities and several basic movements. All subjects were asked to evaluate their own vividness of imagery on Betts's and VMI test with five-point rating scales. The following results were mainly found out: 1. Nine factors were extracted from the factor analysis including both Betts's and VMI test items. Each of these nine factors was correlated with either Betts's test items or VMI test items, but not with the both. It seems that the vividness of motor imagery might be independent from the vividness of general imagery at some degrees. Therefore, the vividness of motor imagery should be measured at the different view point from the vividness of general imagery. 2. In the factor analysis of VMI test, five factors were reasonably interpreted. The first factor was named as visual Imagery of General Movements. This factor seems to indicate visual dominance of motor imagery. After that, the following factors were detected in order: Compound Sense-Imagery of Throwing and Hitting, Compound Sense-Imagery of Gymnastic Movements, Compound Sense-Imagery of Tapping, and Compound Sense-Imagery of Running. It seems that the first factor might be an imagery of "seeing" and the others might be an imagery of "performing". 3. Concerning the first factor of VMI test, the similar factor structures mentioned in 2 were found out both for PE and Non-PE subjects. The other factors for PE majors, however, were more specific in regard to movements than the ones for Non-PE majors. As to these results, it was indicated that characteristics of the vividnese of motor imagery factors were changeable with athletic experiences and acquisition of motor skills. 4. Non-PE majors showed significantly higher factor scores on Visual imagery of General Movements than PE majors. PE majors showed significantly higher factor scores on Compound Sense-Imagery of Gymnastic Movements and Compound Sense-Imagery of Running than Non-PE majors. Probably, an ability of vividness of motor imagery can be differentially affected by athletic experiences.