We describe a method for generating parametric images of the myocardial metabolic rate of glucose (MMRGlc) with positron emission tomography (PET). The method employs serially acquired images of 2-[18F]fluoro-2-deoxy-D-glucose (FDG) uptake and a Patlak graphical analysis of the image data. The arterial input function is derived from images of the left ventricular blood pool calibrated with 18F-plasma measurements. The approach is computationally fast enough to be used in a clinical environment. The MMRGlc parametric images improve myocardial contrast relative to non-parametric images, especially in studies with poor myocardial uptake of FDG. In addition, MMRGlc parametric images consolidate the large amount of data in a dynamic PET study into a clinically usable image set.