In initial studies using fluorine-18-fluorodeoxyglucose (FDG) in normal fasted subjects, we observed disparities in the regional myocardial accumulation of this tracer. Accordingly, we systematically evaluated regional myocardial FDG accumulation in comparison with regional myocardial perfusion assessed with oxygen-15-water and oxidative metabolism assessed with carbon-11-acetate in nine normal subjects (four studied after a 5-hr fast and five studied both fasted and following glucose loading). Under fasting conditions, myocardial accumulation of FDG in the septum and anterior wall averaged 80% of that in the lateral and posterior walls (p less than 0.03). In contrast, after glucose loading the regional distribution of myocardial FDG accumulation became more homogeneous. Regional myocardial perfusion, oxidative metabolism, and accumulation of carbon-11-acetate were homogeneous under both conditions. Thus, under fasting conditions there are regional variations in myocardial accumulation of FDG, which are visually apparent, are not associated with concomitant changes in oxidative metabolism or perfusion, and cannot be attributed to partial-volume effects. This significant heterogeneity may limit the specificity of PET with FDG for detecting myocardial ischemia in fasting subjects.