Objective: We investigated whether baseline vessel status evaluated by magnetic resonance angiography (MRA) can be the foremost factor to classify acute ischemic stroke patients into subgroups for thrombolytic therapy within 3-6 hours of symptom onset. Methods: Acute ischemic stroke patients beyond 3 hours after symptom onset were examined by stroke magnetic resonance imaging (MRI) (diffusion- and perfusion-weighted imaging, and MRA) before and after thrombolysis treatment within 24-48 hours. Stroke MRI was used to classify acute ischemic stroke patients into subgroups and select optimal patients for thrombolytic treatment. Clinical scores were compared to determine whether there were significant differences among subgroups. Results: The difference in day 90 modified Rankin scale (mRS) between treated salvageable and untreated salvageable patients with recombinant tissue plasminogen activator (rt-PA) was remarkably statistically significant (p = 0.02). Treated salvageable patients had more favorable clinical outcomes as compared with the untreated salvageable patients. Patients who did not have baseline artery occlusion were associated with more favorable clinical outcomes than untreated salvageable patients (p < 0.001). The difference between treated salvageable and patients without artery occlusion in 90 day mRS score was not statistically significant (p = 0.058). Conclusion: Baseline vessel status evaluated by MRA may be used as the first factor ahead of mismatch to categorize acute ischemic stroke patients into subgroups. Patients who do not have initial vessel occlusion may not need thrombolytic therapy. [Neurol Res 2009; 31: 355-361]