Purpose: To discuss the difference in the exercise participation motivation, perceived exercise benefit, perceived exercise barrier, exercise self-efficacy and exercise behavior in terms of gender and the predictions of each variable on the exercise behavior. Methods: The research targets were 146 teaching staff from colleges and universities, which involved 101 male participants with the average age being 47.54 ± 10.06 years and 45 female participants with the average age being 39.20 ± 8.95 years. The research tools used included the scales for exercise participation motivation, exercise self-efficacy, perceived exercise benefit and perceived exercise barrier and the investigation on exercise behavior. The data analysis methods used included descriptive statistics, Pearson's product-moment correlation coefficient, t-test and multiple hierarchical regression analysis. Results: Gender did not make any difference in exercise participation motivation (psychological needs, weight control and disease prevention) and exercise behavior. However, in terms of exercise participation motivation (social needs and combined benefits), perceived exercise benefit and exercise self-efficacy, male participants were higher than female participants. Moreover, for perceived exercise barrier, female participants were higher than male participants. In terms of the predictions for the exercise behavior using exercise participation motivation, exercise self-efficacy, perceived exercise benefit and perceived exercise barrier, only exercise self-efficacy had the prediction ability. Conclusion: The exercise participation motivation, exercise self-efficacy, perceived exercise benefit and perceived exercise barrier for teaching staff from colleges and universities to carry out exercise behavior were discussed and suggestions on the exercise behavior promoting strategies were proposed for people in this group.