The medical community has not yet identified cerebrovascular pathophysiological factors that distinguish patients at high risk for stroke or aid in selecting patients for microvascular cerebral bypass. In this study, we describe the courses of 13 patients, all of whom suffered recurrent episodes of transient cerebral ischemia after previous cerebral infarction. These patients underwent regional cerebral blood flow studies using xenon inhalation with a CO2 challenge before and at various times after extracerebral-to-intracerebral microvascular anastomosis. Collateral circulation was assessed in all patients before surgery using four-vessel cerebral angiography. Patients were followed for a mean of 30 months (range, 1-7 yr) after the anastomosis. Measurements of mean cortical cerebral blood flow, as measured using the initial Slope Index, and CO2 cerebrovascular reactivity of these 13 patients were compared with those in a group of 20 patients designed as controls. Hemispheric cortical blood flow was significantly depressed in these patients before surgery compared with those in the control group (P less than 0.05). After the bypass, the mean resting Initial Slope Index in these patients increased 14% (P = 0.0005). Cerebral blood flow both before and after CO2 inhalation improved significantly in these patients after surgery (P = 0.001). Detectors bordering computed tomographic or magnetic resonance image documented infarctions, identified as peri-infarct regions, and demonstrated significant mean increases in both cerebral blood flow (38.8-43.2 ml/min/100 g, P = 0.05) and CO2 cerebrovascular reactivity in these patients after bypass (1.71 + 1.91% to 4.00 + 2.38% change Initial Slope Index/mm Hg CO2, P = 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)