Objective: Pathological exercise is a dangerous behavior often observed in eating disorders. Data investigating associated characteristics of pathological exercise in men are lacking, despite college men and women being at equally elevated risk for developing eating disorders. Participants: Two hundred and twenty-four college men who exercise regularly completed a series of self-report questionnaires. Methods: Latent profile analysis was used to identify empirically-derived homogenous subgroups of regular exercisers based on severity of other eating disorder attitudes and behaviors. Profiles were also compared on differences in exercise motivation and general psychopathology (i.e., depression, anxiety). Results: Fit indices indicated a three-profile solution. Profiles described an eating psychopathology group, a low psychopathology group, and a high exercise frequency group without eating disorder features. Conclusions: Pathological exercise cannot be identified using exercise frequency alone. Other features like body dissatisfaction and exercise motivation style are relevant in identifying pathological exercise behavior in college men.