To evaluate the comparative abilities of gated single photon emission computed tomography (SPECT) wall thickening, delayed thallium-201 (Tl-201) SPECT, and F-18 fluorodeoxyglucose (FDG) SPECT in detecting myocardial viability, 23 patients with previous myocardial infarction and clinically suspected viability were studied. Each patient had at least 1 extensive fixed perfusion defect on rest/stress technetium-99m sestamibi SPECT. A total of 41 major vascular territories had fixed defects. The mean (+/- 1 SD) left ventricular ejection fraction determined from gated perfusion SPECT was 26% +/- 11%. Wall thickening was assessed in a semiquantitative fashion by the regional increase in myocardial intensity during systole and was considered normal when a > or = 20% increase was observed. Tl-201 SPECT was acquired 4 hours after resting tracer injection was administered. Viability was considered present when regional defect Tl-201 count density, determined by quantitative analysis, was > 20% greater than that on the resting sestamibi scan. FDG SPECT was performed independently with a 10 mCi F-18 FDG dose after oral glucose loading was performed. A camera equipped with ultrahigh energy collimation was used. Quantitative criteria for viability were the same as for Tl-201. In the 23 patients viability within the fixed sestamibi defects was manifest by preserved wall thickening in 8 patients, delayed Tl-201 uptake in 10 patients, and FDG uptake in 18 patients. Nine major vascular territories with fixed defects were judged viable by wall thickening, 11 by Tl-201 SPECT, and 24 by FDG SPECT (P = .0009). We conclude that FDG SPECT demonstrates more evidence of myocardial viability than either gated sestamibi wall thickening or delayed Tl-201 SPECT.