Regional cerebral blood flow and vasomotor reactivity were measured in 33 patients with surgically remediable hemispheric ischemia by the 133Xe inhalation method prior to superficial temporal to middle cerebral artery (STA-MCA) by-pass. Thirteen patients also underwent LCBF and L lambda measurements by the stable xenon CT method for comparison. Twenty-four had proximal occlusion of one or both internal carotid arteries, 9 had intracranial occlusive disease (4 internal carotid, 5 middle cerebral). Measurements were repeated at intervals up to 30 months following surgery and compared to measurements in a similar group (N = 13) treated medically. In the surgically treated group 22 patients had recurrent TIAs, of whom 12 also had minor residual neurological deficits from recent small cerebral infarctions with potential for recovery (RINDs) while the remaining 11 had RINDs without TIAs. After surgery 28 improved with cessation of TIAs and/or neurological recovery, 3 remained unchanged, 2 cases worsened. Compared to age-matched normal hemispheric F1 (gray matter) values, pre-operative F1 values in the STA-MCA group were reduced in both ischemic and opposite hemispheres. Ischemic regions showed imparied vasomotor reactivity to 5% CO2 or 100% O2 inhalation. After surgery, mean hemispheric F1 values increased + 12.8% on the by-pass side and + 10.5% on the contralateral side. Mean F1 increases reached a maximum 3 months after by-pass, most evident in ipsilateral frontal regions (+ 24.2%). Vasomotor reactivity did not significantly improve. Medically treated cases did not show similar F1 increases. Thirteen with carotid occlusive disease (8 with TIAs, 5 with small recent infarcts) underwent CT LCBF and L lambda measurements before and after STA-MCA by-pass. Cases with recent infarcts showed reduced LCBF and L lambda values which increased significantly after STA-MCA by-pass, however the total group operated upon showed only trends for CBF increases, probably due to large standard deviations encountered in serial measurements.