To estimate regional myocardial glucose utilization (rMGU) with positron emission tomography (PET) and 2-[18F]fluoro-2-deoxy-D-glucose (FDG) in humans, we studied a method which simplifies the experimental procedure and is computationally efficient. This imaging approach uses a blood time-activity curve derived from a region of interest (ROI) drawn over dynamic PET images of the left ventricle (LV), and a Patlak graphic analysis. The spillover of radioactivity from the cardiac chambers to the myocardium is automatically removed by this analysis. Estimates of rMGU were obtained from FDG PET cardiac studies of six normal human subjects. Results from this study indicate that the FDG time-activity curve obtained from the LV ROI matched well with the arterial plasma curve. The rMGU obtained by Patlak graphic analysis was in good agreement with direct curve fitting results (r = 0.90). The average standard error of the estimate of the Patlak rMGU was low (3%). These results demonstrate the practical usefulness of a simplified method for the estimation of rMGU in humans by PET. This approach is noninvasive, computationally fast, and highly suited for developing parametric images of myocardial glucose utilization rate.