Using a noncontact displacement transducer, the authors measured protrusion of the feline cortical surface through a standardized craniotomy during acute equi-MAC exposures to enflurane, halothane, and isoflurane. Each agent was studied at 0.5, 1.0, and 1.5 MAC concentrations (plus 75% N2O) without support of blood pressure. A repeat 1.0 MAC exposure was made, during which angiotensin was infused to maintain mean arterial pressure (BP) at approximately 145 mmHg. Normocapnia was maintained during all studies. In the absence of BP support, halothane produced significantly greater protrusion of the brain surface than did equi-MAC concentrations of isoflurane at all levels and greater protrusion than enflurane at 1.0 and 1.5 MAC. Halothane-induced protrusion exceeded that seen during isoflurane administration by a factor of 2.5 at 0.5 MAC (P < 0.05); by 1.7 at 1.0 MAC (P < 0.01); and by 2.2 at 1.5 MAC (P < 0.001) and exceeded that seen during enflurane administration by a factor of 1.6 at 1.0 MAC (P < 0.01) and 2.2 at 1.5 MAC (P < 0.001). When anesthetic-induced differences in BP were eliminated by arterial pressure support, the disparity between the protrusion caused by halothane as compared with that caused by enflurane and isoflurane (1.0 MAC) was exaggerated. At similar BP, (during 1.0 MAC exposure), halothane produced approximately 2.4 times as much protrusion as both enflurane and isoflurane (P < 0.0001). The results indicate that enflurane (1.0 and 1.5 MAC) and isoflurane (all levels) cause markedly less protrusion of the brain into a craniotomy than does halothane. The findings roughly parallel the known effects of these agents on cerebral blood flow and probably reflect differences in anesthetic-induced changes in cerebral blood volume. If applicable to human anesthesia, they suggest that in situations during intracranial surgery where administration of a volatile anesthetic is deemed preferable to the use of an additional fixed agent, that isoflurane may be the volatile agent of choice.