Because the literature suggests that aerobic exercise is associated with physical health and psychological well-being, there is a concern with discovering how to improve adherence to such exercise. There is growing evidence that self-motivation, as measured by the Dishman Self-Motivation Inventory (SMI), is a redictor of adherence to regular exercise. Since self-motivation may be amenable to skills training, it is important to understand what skills constitute self-motivation. A study was conducted to identify correlates of self-motivation in a group of persons enrolled in a marathon clinic. Volunteers (N=220) involved in a 9-month adherence study completed the SMI, Heiby's Frequency of Self-Reinforcement Questionnaire, Wallston's Health Locus of Control Scale, Zung's Self-Rating Depression Scale, Speilberger's State-Trait Anxiety Inventory, and an exercise scale. The exercise scale assessed enjoyability of exercise, perceived ability to exercise, current exercise habits, motivation to exercise, education, and physical characteristics. The results provide concurrent and construct validity support for the SMI. It was found that scores on the SMI were positively related to measures of self-reinforcement and internal locus of health control, and negatively correlated with the adverse emotional states of anxiety and depression. If self-motivation is amenable to skills training, norms for the SMI must be developed so that persons with low self-motivation skills can be identified for training in an attempt to enhance compliance to a prescribed activity. A three-page list of references concludes the document. (NB)