BACKGROUND In patients with known or suspected coronary artery disease, diagnosis and risk stratification can be aided by non-invasive tests for myocardial ischaemia. The main indications for the use of radionuclide perfusion imaging (sestamibi) rather than exercise electrocardiography include complete left bundle branch block, inconclusive stress electrocardiography and an inability to exercise. The published data on myocardial perfusion imaging is limited to eight studies of only 628 patients with widely varying ranges of sensitivities and specificities and the major problem being the bias created when patients with positive test results are referred far more frequently for coronary angiography than those with negative results. Furthermore, certain categories of patients are postulated to have a high number of false positives, especially obese patients and women, but this hypothesis has not been tested. RESULTS A retrospective descriptional study on patients with positive sestamibi scans, who had coronary angiography within six months of each other, was performed. The study population consisted of 132 patients, 53.8% of whom were male and 44.7% had a body mass index of less than 28. The overall positive predictive value of the sestamibi scan in predicting significant coronary angiographic stenosis was 63.6%. This value increased significantly to 80.3% in males (p = 0.000). The positive predictive value did not differ in the different age categories, nor did it change with the different indications for sestamibi scanning. CONCLUSIONS The sestamibi scan remains a useful screening test for coronary artery disease in patients who are exercise intolerant or those with inconclusive stress electrocardiography. The test is highly predictive in males rather than females.