Technetium-99m methoxyisobutyl isonitrile (technetium-99m sestamibi [MIBI]) is distributed in the myocardium according to blood flow. Reports comparing stress rest sestamibi protocols with reinjection thallium or resting fluorodeoxyglucose (FDG), or both, in patients with coronary artery disease have shown appreciable discordance regarding myocardial viability in these settings. We performed this analysis with regard to regional comparisons within discordant segments and made comparisons in a subset of patients who underwent revascularization. Thirty-seven patients with coronary artery disease had single-photon emission computed tomography MIBI, N-13 ammonia/18FDG positron emission tomography (PET), and radionuclide ventriculography performed at rest. One hundred two segments were viable and 29 were nonviable by both MIBI and FDG. The concordance was 71%. In MIBI nonviable/FDG viable segments, most of the discordance was in the inferior wall. In MIBI nonviable discordant segments, FDG accurately predicted an increase in percent regional ejection fraction (preoperative 36% [+/- 5 SE] to postoperative 48% [+/- 5.5 SE][p < 0.0006]). MIBI underestimates myocardial viability as assessed by PET. Seventy-one percent of myocardial segments were concordant by both quantitative sestamibi single-photon emission computed tomography and FDG PET. Discordance in MIBI nonviable segments was predominantly in the inferior wall. PET can be helpful in detecting myocardial viability in patients suspected of having had MIBI nonviability in the inferior wall.