To test the effect of verapamil and diltiazem in acute stroke, three groups of mongrel cats of either sex underwent occlusion of the middle cerebral artery (MCA) via a transorbital approach under ketamine anesthesia. The first group served as controls, the second received an intravenous infusion of verapamil (0.1 microgram/kg/min), and the third received an intravenous infusion of diltiazem (0.1 to 1.0 microgram/kg/min). All drug infusions began 2 hours before MCA occlusion and continued for the remainder of the experiment. Before and for up to 24 hours after MCA occlusion, regional cerebral blood flow (rCBF), somatosensory evoked potentials (SSEP's), arterial blood gases, blood pressure, temperature, and hematocrit were measured at least every 2 hours. At the experiment's end, brains were perfused with India ink, removed, sliced, photographed for determination of nonperfused brain area, and weighed, dried, and reweighed for H2O content determination. In these studies, verapamil was associated with worsening of rCBF in ischemic regions and inappropriate increases in rCBF in nonischemic regions, indicating intracerebral steal. Diltiazem increased rCBF in marginally ischemic regions. Changes in SSEP's paralleled blood flow changes, with verapamil decreasing amplitude and conduction velocity while diltiazem slightly improved conduction in the ischemic brain. Verapamil increased the area of nonperfused brain and the content of cerebral H2O. Diltiazem-treated animals had decreased cerebral H2O content, but had a marked increase in the area of nonperfused brain, a finding associated with the high incidence of transtentorial herniation in the diltiazem-treated animals. These findings agree with in vitro studies demonstrating high sensitivity of cerebral blood vessels to calcium channel blockers. These studies further support the notion that calcium channel blockers probably affect several different classes of calcium channels, at different brain sites.