The carotid injection technique, used previously to quantitate the kinetics of blood-brain barrier transport of metabolic substrates, may be modified to analyze the rate of cerebral glucose utilization. A 0.2-ml solution of [14C]glucose (GF) and [3H]methylglucose (M), an internal reference, is rapidly injected into the carotid artery, followed by microwave fixation of brain at various times up to 4 min after injection. The brain radioactivity is separated into a fraction containing neutral hexoses (GF and M) and a fraction containing metabolites of glucose. The GF/M ratio is related to the rate constant (k3) of brain glucose utilization by the simple, linear equation: 1n(GF/M) = ln(GF0/0)--k3t, where GF0/M0 = the brain uptake index of glucose, relative to methylglucose, at 5--15 s after injection, and t = the time after carotid injection, e.g., 1--4 min. It is assumed that (a) the rate of influx due to recirculation of label is minimal during the 4-min circulation period; and (b) the rate constants of glucose efflux (k2) and methylglucose efflux (k2*) are identical. Independent estimates of k2 and k2* showed these parameters to be identical: k2 = 0.14 +/- 0.08 min-1; k2* = 0.14 +/- 0.02 min-1. A logarithmic plot of GF/M ratios versus time was linear (r = 0.99), and was described by the slope k3 = 0.21 +/- 0.02 min-1. Assuming glucose is uniformly distributed in brain, then the glycolytic rate = k3 x brain glucose = (0.21 min-1) (2.6 mumol g-1) = 0.55 mumol min-1 g-1 for the cortex of the barbiturate-anesthetized rat. These studies provide the basis for a simple method of measurement of regional brain glycolysis that does not require either the use of correction factors, e.g., the lumped constant, or the use of differentially labeled glucose.