The initial validation of the brief inhalation method to measure cerebral oxygen extraction fraction (OEF) with positron emission tomography (PET) was performed in non-human primates with predominantly normal cerebral oxygen metabolism (CMRO2). Sensitivity analysis by computer simulation, however, indicated that this method may be subject to increasing error as CMRO2 decreases. Accuracy of the method under pathologic conditions of reduced CMRO2 has not been determined. Since reduced CMRO2 values are observed frequently in newborn infants and in regions of ischemia and infarction in adults, we determined the accuracy of the brief inhalation method in non-human primates by comparing OEF measured with PET to OEF measured by arteriovenous oxygen difference (A-VO2) under pathologic conditions of reduced CMRO2 (0.27-2.68 ml 100g-1 min-1). A regression equation of OEF (PET) = 1.07 x OEF (A-VO2) + 0.017 (r = 0.99, n = 12) was obtained. The absolute error in oxygen extraction measured with PET was small (mean 0.03 +/- 0.04, range -0.03 to 0.12) and was independent of cerebral blood flow, cerebral blood volume, CMRO2, or OEF. The percent error was higher (19 +/- 37), particularly when OEF is below 0.15. These data indicate that the brief inhalation method can be used for measurement of cerebral oxygen extraction and cerebral oxygen metabolism under pathologic conditions of reduced cerebral oxygen metabolism, with these limitations borne in mind.