Objectives The objectives of this study were to (i) investigate directional influences of self-efficacy, outcome satisfaction, and attendance during an exercise programme and (ii) examine the predictive capabilities of self-efficacy and outcome satisfaction on participant dropout. Methods Adults aged 35-65 years were recruited to a 12-month exercise programme. Self-efficacy was collected at baseline, three, six, nine, and 12 months, and outcome satisfaction at the same time-points except baseline. Cross-lagged panel analyses and logistic regression analyses were conducted to examine the primary and secondary objectives, respectively. Results Coping and scheduling self-efficacy had stronger reciprocal relationships with outcome satisfaction than task self-efficacy, although the strength of these relationships varied across time. Initially, task self-efficacy predicted programme attendance. Midway through the programme, outcome satisfaction predicted attendance, and by the end of the programme, coping self-efficacy was the strongest predictor of attendance. Self-efficacy and outcome satisfaction did not predict programme dropout. Conclusions The results support the reciprocal relationship between outcome satisfaction and scheduling and coping self-efficacy and highlight the importance of sequentially targeting different beliefs and skills to facilitate successful initiation and maintenance of exercise behaviour. Statement of contribution What is already known on this subject? Self-efficacy and outcome satisfaction are related to exercise behaviour. The relationship between self-efficacy and outcome satisfaction are typically considered unidirectional. The longitudinal relationships among self-efficacy, outcome satisfaction, and physical activity participation are unclear. What does this study add? Outcome satisfaction had reciprocal relationships with both coping and scheduling self-efficacy for exercise. The importance of self-efficacy and outcome satisfaction to exercise attendance changes over time. Task self-efficacy was the strongest predictor of initial programme attendance. Midway through the programme, outcome satisfaction and task self-efficacy were the strongest predictors of attendance, and by the end of the programme, coping self-efficacy was the strongest predictor.
