OBJECTIVES: To study the clinimetric properties of the Dijon Physical Activity Score (PAS) in patients with coronary artery disease (CAD). PATIENTS: Two populations of patients with CAD: one group of stabilized patients from the RICO county-wide monitoring program and one group in the initial phase of a cardiovascular rehabilitation program (CVR group). METHODS: The patients carried out a maximal effort test on a cycle ergometer, plus two walking tests (a six-minute walk test and a 200 m fast walk test). They completed the Dijon PAS questionnaire on two occasions at an interval of 10 days. The reproducibility of the score and the latter's correlations with physical parameters were analyzed. RESULTS: Sixty-seven subjects were included and 52 answered the questionnaire both times. The average time spent answering the questionnaire was 173+/-37 seconds and reproducibility was satisfactory in the RICO group only. In this group, there were significant correlations between total score and maximal power during the effort test (r=0.41; P<0.05) and between the "sports/leisure activities" sub-score and maximal power (r=0.57; P<0.01). No correlations were found in the CVR group. CONCLUSION: The Dijon PAS is a simple, generic, reproducible and reliable score for measuring physical activity in patients with stable coronary artery disease but, because of the conjunction of confounding factors, it is not suitable for subjects who experienced a recent acute cardiac event. It could thus be used in epidemiological studies to determine the impact of a sedentary lifestyle and the efficacy of methods intended to counter sedentariness and to help design personalized secondary prevention programs.