A brain efflux index method has been developed to characterize an efflux transport system for substrates from the cerebrum to the circulating blood across the blood-brain barrier. The brain efflux index value is defined as the relative amount of test drug effluxed from cerebrum compared with that of a reference compound, [14C]carboxylinulin, which has limited blood-brain barrier permeability. Microinjection of 0.2, 0.5 or 1.0 microliter into the parietal cortex area 2 region was found to be an appropriate procedure for obtaining a high recovery and for restricting the test drug and reference compound to the ipsilateral cerebrum. No significant increase in influx clearance of [14C]carboxylinulin into the brain was observed for the ipsilateral cerebrum after the sham microinjection compared with the contralateral cerebrum, which demonstrated that only limited physical damage is caused. Microinjection of [3H]H2O into the cerebrum devoid of cerebral blood flow yielded no elimination from the brain. Analysis based on a pharmacokinetic model demonstrated that the elimination of a highly permeable compound [3H]H2O from the brain was governed by cerebral blood flow. The apparent elimination rate constant (Kel) of [3H]3-O-methyl-D-glucose from the brain was determined as 0.129 +/- 0.014 (min-1) and was reduced significantly by a preinjection of excess cold 3-O-methyl-D-glucose and phloridzin, whereas no significant elimination was observed for L[3H]glucose. The efflux clearance of 3-O-methyl-D-glucose was calculated from the Kel value and the compound's distribution volume, the value being close to that of the influx clearance. These results demonstrate that the brain efflux index method is a useful technique for analyzing an efflux process from the brain across the blood-brain barrier involving a carrier-mediated transport system.