The methodology and questions were selected following two major perspectives concerning how persons change health practices guide the present study: 1) the transtheoretical model of behavior change (Prochaska & DiClemente, 1983) and Bandura's social cognitive theory of social learning theory (Bandura, 1977; 1989). The transtheoretical model states that persons change their health behaviors by proceeding through a process of stages: precontemplation (no thought of changing), contemplation (aware of need to change), preparation (takes small steps), action (modifies behavior), and behavior maintenance (incorporates change into routine). The model allows us to understand when particular shifts in attitudes, intentions, and behaviors occur. Two studies have reported that pre-treatment stages of change according to this model are associated with movement to active participation in exercise activities (Barke & Nicholas, 1990; Marcus, Rakowski, & Rossi, 1992). This model includes the concept of decisional balance, which refers to one's evaluation of the personal gains and losses with changing behavior. Marcus and colleagues found that individuals are more likely to exercise if the perceived gains outweighed the perceived losses. However, no study has examined the applicability of this model to persons with MR. Bandura's social cognitive theory posits that behavior change is a function of setting goals based on outcome expectations associated with the behavior change, the tasks required to achieve those goals, and self-efficacy expectations for achieving the goals (Bandura, 1977; 1989). Thus, individuals are more likely to change their exercise behaviors if they believe that: a) their current lifestyles pose threats to personally valued goals; b) exercise behaviors will help reduce the threat (outcome expectations); and c) they are personally capable of adopting the new behaviors (self-efficacy expectations). This theory has been applied widely to the study of health behaviors. Self-efficacy has been shown to be a major predictor of adherence to preventive health programs (O'Leary, 1985), and to exercise involvement (McAuley, Lox, & Duncan, 1993). Bandura also emphasizes the informative and motivational role of reinforcement and observational learning through modeling the behavior of others. Hence, environmental cues, including support from others, play an important role. Berkman (1995) notes that interventions aimed at restructuring naturally occurring networks will be more effective than those that rely on short-term constructed support groups. A few exploratory studies suggest that social support is essential in promoting positive health behaviors for adults with MR (Fox, et al., 1985; Fujiura et al., 1997). The present study is designed to investigate the applicability of both the transtheoretical model in delineating stages of behavioral change and the social cognitive model in explaining the factors predicting long-term adherence to an exercise program among adults with a life-long cognitive disability, such as adults with MR.