The effects of patient habitus (e.g., breast attenuation in women and diaphragmatic attenuation in men) have long been recognized as factors that reduce the accuracy of myocardial perfusion imaging. Although it has long been assumed that patient obesity effects accuracy, this has never been formally evaluated. We studied the effects of patient obesity, defined as a body mass index (BMI) > or = 30, on 607 patients who underwent exercise thallium-201 single-photon emission computed tomography (SPECT). Because the effects of obesity are most likely mediated through increased photon attenuation and scatter, we also evaluated the effects of other markers of patient size: body surface area (BSA) and patient weight. Accuracy was determined by performing quantitative analysis and measuring the area under the receiver operating characteristic curve (AUC). Obesity was associated with significantly lower accuracy (AUC 0.86 +/- 0.03 vs 0.92 +/- 0.02, p <0.05) despite similar estimates of maximal coronary blood flow (as estimated by heart rate and rate-pressure product at peak exercise) and severity of coronary disease. There were no significant differences attributable to either patient weight or BSA. Weight and BSA correlated significantly with left ventricular chamber size whereas BMI did not. We conclude that the accuracy of quantitative SPECT thallium-201 is significantly reduced by patient obesity and that although BSA and weight are also associated with increased attenuation, they have no effect on accuracy, which is most likely due to the compensating effects of increased chamber size.