Given that ~50% of all exercise intenders will fall into the intention-behavior gap (i.e., a situation where people fail to act on their intentions), it is necessary to identify the constructs and/or theories that can explain the discord between intention and behavior (i.e., the intention-behavior gap). For this purpose, the present research was conducted through two studies that were designed to test the efficacy of causal attributions as a means to reduce the intention-behavior discord. The first study collected information from 952 individuals on their exercise behavior and their associated causal attributions over a six-week period. The findings from this study included: (1) those individuals who fell into the intention-behavior gap made self-serving attributions for their exercise failure; (2) Weiner’s model accurately predicted several of the affective and cognitive responses to exercise behavior for the sample of exercise intenders; and (3) causal attributions were not found to be effective moderators of the intention-behavior relationship. The second study was an experiment that tested whether an attribution retraining intervention could improve exercise behavior for a sample of sedentary, exercise intenders (n=200). Results of this study were mixed as the intervention appeared to have been able to modify one of the targeted attributional dimensions (control), but the effect was not strong enough to change the exercise behavior of the participants in the experimental group. It is suggested that attributions may not be able to reduce the gap because they represent conscious deliberations of the behavior, while sustained exercise is based on nonconscious processing of relevant information to make exercise an automatic behavior.