An autoradiographic sequential double-label variant of the deoxyglucose method for measurement of local glucose utilization has been developed. This technique takes advantage of the short half-life of the positron emitter, 18F (t 1/2 = 110 min) relative to that of 14C. Sequential injection of [ 18F]fluorodeoxyglucose (FDG) and 14C-labeled FDG allows the production of two separable autoradiograms, each of which represents the same 20-μm brain slice, but under potentially different cerebral metabolic states. We have used this technique to demonstrate that ibotenic acid-lesioned rat striatum is selectively refractory to the depressing effects of barbiturate anesthesia upon brain glucose utilization. The described method has applicability to the analysis of small changes in regional cerebral metabolism in localized brain regions and represents a solution to the problem of intersubject variability inherent in conventional approaches to the deoxyglucose method.