A method was developed to measure simultaneously the rate constants for glucose influx and glucose efflux, and the Michaelis-Menten constant (KM) and maximal velocity (Vmax) for glucose transport across the blood-brain barrier (BBB) in any selected brain area. Moreover, on the basis of a mathematical model, the local perfusion rate (LPR) and local unidirectional glucose transport rate (LUGTR) are calculated in terms of parameters of the time-activity curves registered over different brain regions; 11C-methyl-D-glucose (CMG) is used as an indicator. The transaxial distribution of activity in the organism is registered using dynamic positron-emission tomography (dPET). The method was used in 4 normal subjects and 50 patients with ischemic brain disease. In normals, the rate constant for CMG efflux was found to be 0.25 +/- 0.04 min-1 in the cortex and 0.12 +/- 0.02 min-1 in white matter. In the cortex, the KM was found to be 6.42 mumol/g and the Vmax was 2.46 mumol/g per minute. The LUGTR ranged from 0.43 to 0.6 mumol/g per minute in the cortex, and from 0.09 to 0.12 mumol/g per minute in white matter. The LPR was calculated to be 0.80-0.98 ml/g per minute for the cortex and 0.2-0.4 ml/g per minute for white matter. In patients with stroke, the ischemic defects appeared to be larger in CMG scans than in computed x-ray tomography (CT) scans. Prolonged reversible ischemic neurological deficit was associated with a significant fall in the LUGTR but no change in the LPR in the corresponding cerebral cortex. Normal LUGTR and significantly decreased LPR were registered in a patient with progressive occlusion of the middle cerebral artery. In a patient with transient ischemic attacks, a slightly reduced LPR and a disproportionally reduced LUGTR were observed before operation. After extra- and intracranial bypass surgery, the LPR became normal, whereas the LUGTR increased but did not achieve normal values.