Background. The majority of research aiming to predict physical activity (PA) has applied social cognitive models. Although these models have been effective in identifying key correlates and predictors, such approaches are limited as they do not account for post-intention, implicit, and dispositional type constructs. The Integrated Behavior Change (IBC) model incorporates tenets from previous theories found to predict PA and includes pathways relevant to implicit processes. Previous findings have supported predictive validity of the IBC model for nutrition choices and sun safety behaviour, but it has not been tested in PA. The purpose of this study was to test the IBC model in a sample of gym/recreation center attendees. Methods. Participants (N = 161) were a sample of adults randomly selected from nine gym and recreation centres who completed baseline and follow-up questionnaires after eight weeks. Hypotheses of the IBC were tested using structural equation modeling. Prior to analysis, all constructs were converted to standardized residual scores to represent change in constructs over eight weeks. Results. Intention x planning (β = .22, p = .004) and habit (β = .32, p < .001) were found to predict PA directly. Intrinsic reward predicted habit (β = .20, p = .025) and intention x planning (β = .33, p = .001). We also found a significant total indirect effect of intrinsic reward on behavior (β = .137, p = .026). Finally, consistency was found to predict habit (β = .30, p = .023). The model explained 21% of the variance in behavior. Discussion. This is the first study to test the IBC model in a PA context. Results support previous research adopting dual process models to predicting PA and highlights the importance of nonconscious processes in predicting behaviour change. Future research should extend study duration and assess behavior using objective measures.