We have evaluated the impact of increased body mass on the quality of myocardial perfusion imaging using a latest-generation g-camera with cadmium-zinc-telluride semiconductor detectors in patients with high ($40 kg/m 2 ) or very high ($45 kg/m 2 ) body mass index (BMI). Methods: We enrolled 81 patients, including 18 with no obesity (BMI , 30 kg/m 2 ), 17 in World Health Organization obese class I (BMI, 30-34.9 kg/m 2 ), 15 in class II (BMI, 35-39.9 kg/m 2 ), and 31 in class III (BMI $ 40 kg/m 2 ), including 15 with BMI $ 45 kg/m 2 . Image quality was scored as poor (1), moderate (2), good (3), or excellent (4). Patients with BMI $ 45 kg/m 2 and nondiagnostic image quality (#2) were rescanned after repositioning to better center the heart in the field of view. Receiver-operating-curve analysis was applied to determine the BMI cutoff required to obtain diagnostic image quality ($3). Results: Receiver-operating-curve analysis resulted in a cutoff BMI of 39 kg/m 2 (P , 0.001) for diagnostic image quality. In patients with BMI $ 40 kg/m 2 , image quality was nondiagnostic in 81%; after CT-based attenuation correction this decreased to 55%. Repositioning further improved image quality. Rescanning on a conventional SPECT camera resulted in diagnostic image quality in all patients with BMI $ 45 kg/m 2 . Conclusion: Patients with BMI $ 40 kg/m 2 should be scheduled for myocardial perfusion imaging on a conventional SPECT camera, as it is difficult to obtain diagnostic image quality on a cadmium-zinc-telluride camera.