This study described the practice profile of an open access exercise stress test (EST) service to the primary care physicians at a teaching hospital in 2000. We performed a retrospective review of all ESTs ordered and conducted by the primary care physicians. A total of 145 ESTs were conducted, of which 80.7% were referred for assessment of chest pain. Proportions of positive, negative, uninterpretable and inconclusive ESTs were: 22.1%, 52.8%, 18.1% and 6.9%. Typical chest pain was independently associated with a positive EST in this study (p = 0.008, OR 5.50, 95% CI 1.56-19.37). Although referral to the open access EST service seemed appropriate, there is a need to reduce the number of uninterpretable and inconclusive results.