The effect of acute cerebral occlusion on the distribution of cerebral perivascular nerves containing catecholamine, neuropeptide Y, and vasoactive intestinal peptide was studied in the three commonly used rat models of cerebral ischemia: Unilateral permanent middle cerebral artery (MCA) occlusion induced with an intraluminal thread technique; unilateral MCA occlusion produced by extraluminal electrocoagulation of the MCA; and transient arterial occlusion of the whole brain induced extracranially by the four-vessel clasp occlusion method for 30 minutes. Animals were sacrificed 3 days after occlusion and the distribution of the perivascular nerves of the MCA studied. Intraluminally occluded MCAs showed a similar distribution of perivascular nerves to those of contralateral and sham-operated MCAs. Extraluminally occluded MCAs demonstrated a marked decrease in perivascular nerves containing catecholamine and peptides while the contralateral MCAs showed normal distribution of the nerves. Extracranial occlusion caused no discernible change in the distribution of perivascular nerves in occluded and sham-operated animals. This study indicates that the different methods of cerebral arterial occlusion have variable effects on the perivascular innervation. Arterial occlusion induced by intraluminal or transient extracranial procedures does not impair cerebral perivascular innervation at least up to 3 days post-occlusion. In contrast, cerebral arterial occlusion by extraluminal electrocoagulation diminishes the perivascular nerves around the occluded cerebral artery.