The accumulation of 2-deoxy-D-glucose-6-phosphate (2DG6P), detected using 31P NMR spectroscopy, has been used as a measure of the rate of glucose uptake, yet the accuracy of this measurement has not been verified. In this study, isolated rat hearts were perfused with different substrates or isoproterenol for 30 min before measurement of either 2DG6P accumulation or [2-3H]glucose uptake, without and with insulin. Basal contractile function and metabolite concentrations were the same for all hearts. The basal rates of 2DG6P accumulation differed significantly, depending on the preceding perfusion protocol, and were 38-60% of the [2-3H]glucose uptake rates, whereas insulin-stimulated 2DG6P accumulation was the same or 71% higher than the [2-3H]glucose uptake rates. Therefore the ratio of 2DG6P accumulation/[2-3H]glucose uptake rates varied from 0.38 to 1.71, depending on the prior perfusion conditions or the presence of insulin. The rates of 2DG6P hydrolysis were found to be proportional to the intracellular 2DG6P concentrations, with a K(m) of 17.5mM and V(max) of 1.4 micromol/g dry weight/min. We conclude that the rates of 2DG6P accumulation do not accurately reflect glucose uptake rates under all physiological conditions in the isolated heart and should be used with caution.