Using rats in which incomplete cerebral ischemia was induced, the authors evaluated the effects of halothane (H) and isoflurane (I) on neurologic outcome compared to nitrous oxide (N2O) controls. Incomplete cerebral ischemia was produced by right carotid artery occlusion combined with hemorrhagic hypotension. Neurologic outcome was evaluated using a graded deficit score from 0 to 5 (0 = normal, 5 = death associated with stroke). Two levels of cerebral ischemia were tested. At moderate ischemia with hypotension of 30 mmHg, an FIO2 of 0.3, and ischemic periods of 30 or 45 min, N2O produced a deficit of 4.7-5.0 and a mortality rate of 90-100%. In contrast, halothane (1 MAC) and isoflurane (1 MAC) resulted in similar deficit scores (H = 1.1-1.8, I = 1.4-1.6) and mortality rates (H = 17-30%, I = 17-20%). Cerebral blood flow (CBF) measured with radioactive microspheres showed a 60-65% decrease in the ischemic hemisphere at this level of hypotension. With severe ischemia with hypotension = 25 mmHg, FIO2 = 0.2, and a 30-min period of ischemia, deficit scores increased to 3.0 and 3.9 with 1 MAC halothane and 1 MAC isoflurane, respectively. Mortality rates also increased to 40% with halothane and 70% with isoflurane. Increasing the concentration of halothane or isoflurane to 2 MAC did not significantly improve outcome. Brain histology demonstrated extensive neuronal damage in striatal, hippocampal, and neocortical regions of N2O control treated rats, and less damage with little difference between H- and I-treated rats at each level of ischemia.(ABSTRACT TRUNCATED AT 250 WORDS)