Background: This study evaluated the usefulness of MR angiography (MRA)diffusion-weighted imaging (DWI) mismatch in neuroendovascular therapy over 3 h after onset of acute cerebral infarction. Methods: The subjects were 14 cases ( age, 73 +/- 8.4 years) who had an anterior circulation deficit on DWI/ MRA on arrival and underwent neuroendovascular therapy over 3 h after onset. MRA-DWI mismatch ( MDM) (+) was defined as 'major artery lesion (+) and diffusion-weighted image-Alberta Stroke Program Early CT Score ( DWI-ASPECTS) 6'; MDM (-) was defined as 'major artery lesion (+) and DWI-ASPECTS < 6'. Results: Reperfusion was achieved in nine of 14 patients ( 64%) undergoing neuroendovascular therapy. Within the reperfusion group, in the five MDM (+) patients and the four MDM (-) patients, the outcome was a favorable clinical response in the MDM (+) group. The modified Rankin Scale ( mRS) scores after 90 days were 0-2 in 3 ( 60%) and 3-6 in 2 ( 40%) of the MDM (+) group patients and 0-2 in 0 ( 0%) and 3-6 in 4 ( 100%) of the MDM (-) group patients. In the MDM (+) group, a good outcome was achieved. However, the number of cases was small, so this was not a significant difference. Within the non-reperfusion group, in the three MDM (+) patients and the two MDM (-) patients, the mRS scores after 90 days were 0-2 in 1 ( 33%) and 3-6 in 2 ( 67%) of the MDM (+) group patients and 0-2 in 0 ( 0%) and 3-6 in 2 ( 100%) of the MDM (-) group patients. In both groups, the outcome was poor. Conclusions: With neuroendovascular therapy, a good outcome with reperfusion was achieved in the MDM (+) group compared to the MDM (-) group. This suggests that the presence or absence of MDM may be useful in determining prognosis after reperfusion.