Background: The IMS-III trial was terminated early for futility after interim analysis showed no difference between intraarterial (IA) and intravenous (IV) treatment arms. This analysis attempts to understand the occurrence and predictors of futile recanalization in the IMS-III. Materials/methods: Data of IMS-III patients treated with IA intervention was analyzed. Patients with near complete or complete recanalization (thrombolysis in cerebral infarction [TICI] gradesII bandIII) were selected and according to functional outcome at 3 months into “useful recanalization” -modified Rankin scale (mRS) 0-2 and “futile recanalization” -mRS 3-6. Subjects with symptomatic intracerebral hemorrhage (ICH), posterior circulation occlusion, or missed data were excluded. Univariate analysis was performed to examine differences between two groups. Logistic regression was performed to identify predictors of futility. Variables selected for logistic regression are those significant in univariate analysis. Results: Complete or near complete recanalization was observed in 145 of 434 patients treated with IA. Only 118 met the criteria for this analysis. Of those, 71 (60[percnt]) patients had useful recanalization and 47 (40[percnt]) had futile recanalization. Compared to patients with useful recanalization, patients with futile recanalization had higher incidence of diabetes mellitus, higher median baseline NIHSS scores (19 vs 15, p<0.001), higher median ASPECTS; 8 vs 7, longer time from onset to IV-tPA (131 vs 116 min, p=0.026), longer time from onset to IA therapy (267vs 240 min, p=0.010), longer hospital stay (13±9vs 6±4 days, p<0.001), and a non-significant trend towards higher mortality rate (1.4[percnt] vs 10.6[percnt], p=0.07). In logistic regression analysis,the only independent predictor of futile recanalization was baseline NIHSS≥20 (odds ratio, 4.1; 95[percnt] confidence interval, 1.1-16.4, p=0.043), while female gender, time to IV-tPA, and ASPECTS showed non significant trend. Conclusions: Futile recanalization is relatively common following IA treatment in IMS-III trial. Important clinical factors can predict this phenomenon, which causes high resources utilization without functional benefit.