The main aim of this study was to investigate if exercisers’ personality characteristics were associated with exercise dependence. Speciﬁcally, the purpose was to examine if anxiety, obsessive passion, and physical appearance orientation were associated to an increased risk for exercise dependence. Participants were 330 exercisers from exercise groups, sport clubs and university sport science classes in the southwest of Sweden. Data were analysed using CHAID (Chi-squared Automatic Interaction Detection) analysis. The CHAID analysis indicated that anxiety was the main predictor of exercise dependence. More speciﬁcally, 12.7% more exercisers who experienced high levels of anxiety symptoms (i.e. scores above 6), were, in comparison to the exercises experiencing low levels of anxiety, classiﬁed as “at risk for exercise dependence”. For exercisers that reported low levels of anxiety symptoms (i.e. scores below 7), obsessive passion for exercise was a positive statistically signiﬁcant predictor (absolute risk difference=8.6%). Overall, the results highlight anxiety as a main risk factor behind exercise dependence. Also, the risk of exercise dependence may increase either from obsessive passion or as a coping strategy for anxiety. Furthermore, results may illustrate two types of exercise dependence; “primary” exercise dependence driven mainly by an obsessive passion for exercise and “secondary” exercise dependence where exercise function as a strategy to cope with anxiety.