The research purposes were to examine the factorial and convergent validity, internal consistency, and test-retest reliability of the Exercise Dependence Scale (EDS). Two separate studies, containing a total of 1,263 college students, were undertaken to accomplish these purposes. Participants completed the EDS and measures of exercise behavior and perfectionism. Confirmatory factor analysis in Study 1 reduced the scale from 28 to 21 items (i.e., 3 items per subscale) and it supported a correlated seven-factor model in Study 2 (Tucker-Lewis Index = 0.96, comparative fit index = 0.97, root mean square error of approximation = 0.05, average absolute standardized residual = 0.02). Adequate internal consistency and test-rest reliability for the scale was evidenced. Consistent with the hypotheses and the EDS classifications for exercise dependence symptoms, individuals at risk for exercise dependence reported more exercise behavior and perfectionistic tendencies, compared to the nondependent-symptomatic and nondependent-asymptomatic groups. These studies provide evidence for the reliability and validity of the EDS. Research in more diverse populations is needed to further examine the general applicability of the EDS-Revised. (Contains 5 tables.)