Background The benefits of supervised physical activity programs in cardiac rehabilitation have been amply demonstrated, but the quantity of physical activity often declines quickly once supervision ends. This trial assesses the effectiveness of an experimental intervention drawing on habit formation theory to maintain physical activity. Methods Cardiovascular patients (N = 47) were randomly assigned to one of two groups. The first group participated in two supervised physical activity (SPA) sessions per week for 20 weeks. The second group was offered a progressively autonomous physical activity (PAPA) program as follows: the same supervised program as the SPA group for 10 weeks and then a further 10 weeks with one supervised session replaced by a strategy to build and sustain the habit of autonomous physical activity. The International Physical Activity Questionnaire (IPAQ; Craig et al. Med Sci Sports Exerc 35(8):1381–1395, 2003) was used to measure the quantity of physical activity, which was the primary outcome. The number of participants was limited, and we thus took multiple IPAQ measurements (at 0, 5, 7, 9 and 12 months after the start of the intervention) and used a mixed model for analysis. Physical condition, automaticity of the physical activity behavior, motivation, and quality of life were examined for changes. Results No significant between-group differences were noted for physical activity behaviors after the program, physical condition, motivation, or behavioral automaticity. The PAPA group nevertheless completed more PA sessions during the intervention, and their quality of life was significantly higher than that of the SPA group at 12 months. Conclusion Although the number of supervised sessions was lower, the progressively autonomous PA program resulted in the same or even higher positive outcomes than the fully supervised PA program.