The fluorodeoxyglucose (FDG) method for the measurement of local cerebral metabolic rate of glucose (LCMRGlc) employs typical values of the FDG transport rate constants that have been obtained by kinetic measurements on an appropriate control group. Discrepancies between the true values of the rate constants in tissue and the typical values used in the operational equation of the FDG method will introduce error in the estimate of LCMRGlc. Computer simulations were used to evaluate the accuracy of the FDG method in cases where (1) the tissue LCMRGlc deviates greatly from the normal values (e.g., stroke) or (2) the tissue LCMRGlc changes during the experiment (e.g., epileptic seizure). The effects of the magnitude and duration of metabolic changes were studied. The results indicate that if tissue LCMRGlc differs greatly from the normal value, the error in the estimated LCMRGlc at a scan time of 60 min is less than 20% of the difference between the true and normal values. In the non-steady-state cases, the estimated LCMRGlc was found to be a weighted average of the metabolic rates during the experiments, with the weightings approximately proportional to the plasma FDG concentration at the corresponding times. For example, if LCMRGlc in tissue was 5 times the normal values for the first 10 min but then returned to normal state, the LCMRGlc measured by the FDG method at a scan time of 60 min would be about only 2-3 times the normal value. The results of this study provide a better understanding of the accuracy of the FDG method under various tissue metabolic conditions and is useful for interpreting metabolic values obtained with the FDG method.