Summary:The cerebral blood volume (CBV) is sensitive to changing hydrostatic pressures. Thus, measurement methods that rely on removing tissue from unfixed brain may lead to underestimates of the CBV due to the loss of blood from the tissue. In situ fixation of tissue before removal may offer improved accuracy. We employed a triple-label method to measure simultaneously whole brain CBF and CBV in halothane-anesthetized Sprague Dawley rats, which were then killed either by focused microwave irradiation (=8 kW of incident power x 770 ms) or by decapitation. CBF was measured with eHJni cotine while the CBV was determined as the sum of the cerebral red cell volume (CRCV-measured with 99mTc_ labeled red cells) and the cerebral plasma volume (CPV measured with [14C]dextran). Animals were studied dur ing hypocarbic (Paco2 = 25 mm Hg), normocarbic, or hypercarbic (P aco2 = 70 mm Hg) conditions. AddedThe cerebral blood volume (CBV) is measured by determining the brain distribution space for nondif fusible red cell and/or plasma markers. Typical trac ers include 51Cr_, 99mTc_, or Cl50-labeled red cells, or 99mTc_ or radioiodine-labeled albumin. In hu mans and in large animals, the brain concentration of such radioactive compounds can be determined in vivo by external detectors (Risberg et aI., 1969; Phelps et aI., 1979; Lammertsma et aI., 1984; Pow ers and Raichle, 1985; Sakai et aI., 1985; Artru, 1988; Archer et aI., 1990). If the arterial concentra- 328 studies were performed to verify that the microwave ir radiation scheme used was capable of fixing previously administered tracers in place, and also halting the entry of tracer given after irradiation. Results indicate that the method of killing had no effects on CBF measurements, as assessed either by absolute values during normocarbia or responsiveness to changing Paco2. However, all three volume measurements made using nondiffusible tracers (CRCV, CPV, and CBV) were significantly lower in ani mals killed by decapitation. Furthermore, CO2 respon siveness for all three variables (as assessed by the slope of the P aco2/volume) was not evident in decapitated ani mals. We conclude that in situ fixation offers significant advantages when examining the cerebral distribution space of nondiffusible tracers.