OBJECTIVE: Internal carotid artery (ICA) balloon test occlusions (BTOs) are performed in the angiography suite to predict whether the patient has adequate collateral circulation to prevent stroke when permanent ICA occlusion (PCO) is required for treatment. Although many criteria have been proposed to facilitate predictions of stroke risk after PCO, no BTO techniques have been subjected to predictive validity testing in outcome studies. We describe a prospective case series study that tests the predictive validity of quantitative cerebral blood flow (CBF) measurements during ICA BTO. METHODS: Thirty-three patients with clinical indications for PCO underwent ICA BTO and then PCO. During BTO, standard neurological examinations, sustained-attention testing, and quantitative CBF measurements were performed. Two scalp scintillation detectors recorded washout data after ipsilateral intracarotid injection of xenon-133 through a port at the tip of the ICA-occluding balloon. Patients were monitored for the outcome measure of ipsilateral stroke for a mean of 34 months. The variables of quantitative CBF values, neurological examination results, sustained-attention test results, age, sex, and side of occlusion were examined with Kaplan-Meier log-rank tests, predictive validity analyses, and logistic regression analyses. RESULTS: CBF of less than 30 ml/100 g/min during BTO was the only variable that predicted stroke after PCO (log rank = 5.87, P = 0.015). The negative and positive predictive values for CBF findings were superior to those for standard neurological examination findings and sustained-attention test results. Age, sex, and side of occlusion did not predict stroke. CONCLUSION: Quantitative CBF testing, via the intracarotid injection technique, during BTO seems to be an important predictor of stroke after PCO.