Rates of glucose utilization (CMRGlc) were determined in some cerebral structures of active warm- and cold-adapted ground squirrels and hibernating ground squirrels with [14C]deoxyglucose (DG) by direct chemical measurement of precursor and products in samples dissected from funnel-frozen brain. The rate of supply relative to demand of glucose and [14C]DG in brain of hibernating animals was similar to or greater than that of controls. [14C]DG cleared from the plasma in hibernators much more slowly than in active animals, and the level of unmetabolized [14C]DG in brain and the integrated specific activity of the precursor pool in plasma exceeded those of the active animals by 4- to 10-fold. At 45 min after an intravenous pulse of [14C]DG, the unmetabolized [14C]DG remaining in the brains of the hibernators accounted for approximately 96% of the total 14C compared with approximately 10-15% in the active animals. The value of lambda, a factor contained in the lumped constant of the operational equation of the [14C]DG method, was estimated for each animal and found to be relatively constant over the sixfold range of glucose levels in the brains of all animals. Calculated CMRGlc in squirrels in deep hibernation was only 1-2% of the values in active animals.