Objective To evaluate the contribution of CT angiography (CTA) in predicting clinical outcome in a broad spectrum of patients presenting with acute neurological deficits suggestive of brain ischemia, to assess its strengths and limitations in this setting, and examine its influence on selection of patients for thrombolytic treatment. Patients and methods Prospective, observational outcome study of 54 consecutive patients with acute neurological deficits suggestive of brain ischemia who received immediate CTA. Clinical outcome was compared for patients presenting with and without arterial occlusion on CTA. Treatment decisions made by a vascular neurologist blinded to CTA results were compared to CTA cognizant treatment. Results For patients presenting with slight to moderate neurological deficits, the sensitivity and specificity for predicting good clinical outcome was 0.62 and 0.79, respectively, using the initial NIH Stroke Scale (NIHSS) score alone, and 0.38 and 0.92 if additionally, CTA showed no occlusion. For patients presenting with more severe deficits, the sensitivity and specificity for predicting poor clinical outcome using the NIHSS score alone was 0.79 and 0.60, compared to 0.67 and 0.92 if CTA showed vessel obstruction. CTA correctly identified six stroke mimickers. Selection of patients for thrombolysis made with knowledge of CTA results were more often conservative, and corresponded to CTA blinded decisions in 42/50 cases (84%, r = 0.72). Conclusions Combining CTA results with the neurological exam allows increased specificity for predicting clinical outcome as compared to predictions based on admission NIH Stroke Scale score alone. Awareness of CTA results was occasionally associated with less aggressive treatment and testing decisions.