Background: A recent randomized controlled trial found that an intervention focused on developing an exercise habit increased weekly minutes of moderate to vigorous physical activity (MVPA) over 8 wk compared to a control group. The purpose of the current study was to test if changes in habit, as well as other behavioral strategy constructs from the Multi-Process Action Control Test, mediated between group condition and MVPA (self-report and accelerometry). Methods: Inactive new gym members (N = 94) were randomized into control or experimental (habit-building) groups. Results: No construct entirely explained mediation condition (experimental and control) and changes in MVPA measured by accelerometry. Self-report MVPA found affective judgments, behavioral regulation, and preparatory habit to be mediated between group (experimental/control conditions) and changes in behavior (β = 0.36, 95% confidence interval [.05–.78]). Conclusions: Self-reported and objectively measured behavior models demonstrated complete and partial mediation, respectively. New gym members could benefit from successful behavioral enactment by developing constructs to support habit formation.