BACKGROUND Because the effect of hypocapnia on distribution of cerebral blood flow during focal cerebral ischemia is controversial, this investigation was performed in rats to determine whether hypocapnia, instituted immediately after the onset of focal cerebral ischemia, produces a favorable redistribution of blood flow (an "inverse steal") toward the ischemic territory. METHODS After surgical preparation during normocapnic isoflurane anesthesia, middle cerebral artery occlusion (MCAO) was performed. Animals were randomized to either immediate institution of hypocapnia (n = 9; PaCO2 23 +/- 2 mmHg) or continued normocapnia (n = 8; PaCO2 40 +/- 2 mmHg). Thirty minutes thereafter, local cerebral blood flow (1-CBF) was determined autoradiographically using 14C-iodoantipyrine. Local cerebral blood flow was determined in four coronal brain sections spanning the distribution of the middle cerebral artery. For the hemisphere ipsilateral to MCAO, the areas of cortex in which CBF fell within three specified CBF ranges (0-6, 6-15, and 15-23 ml/100 g/min) were measured and expressed as a percentage of the total area of cortex in that section. In the hemisphere contralateral to MCAO, to confirm the presence of normal CO2 reactivity in non-ischemic brain in this model, average 1-CBF was determined for the cortex, the subcortex, and the entire hemisphere in each coronal section. RESULTS Hypocapnia resulted in significantly lower 1-CBF in the cortex, subcortex, and entire hemisphere in all coronal sections of brain contralateral to MCAO. In the hemisphere ipsilateral to MCAO, a favorable redistribution of CBF was not observed. For the three more anterior coronal sections (1-3), a significantly larger percentage of the cortex had 1-CBF in the range of 15-23 ml x 100 g-1 x min-1 in the hypocapnia group animals. In sections 2 and 3, significantly larger areas of cortex had 1-CBF in the range of 6-15 ml x 100 g-1 x min-1 in the hypocapnia group than in the normocapnia group. For all sections, there were no significant differences between hypocapnic and normocapnic groups in the area of cortex with 1-CBF in the range of 0-6 ml x 100 g-1 x min-1. CONCLUSIONS The current study does not provide evidence for the occurrence of a hypocapnia-induced inverse steal phenomenon during acute focal cerebral ischemia of 30 min duration in the rat. The data suggest that, rather than reducing the area of the critically ischemic brain, hypocapnia may increase the size of the region at risk. The data do not support the use of hypocapnia as a therapeutic measure to produce a favorable redistribution of CBF during focal cerebral ischemia of acute onset.