Background -Endovascular treatment with mechanical thrombectomy (MT) is beneficial for acute stroke patients suffering a large vessel occlusion, though treatment efficacy is highly time-dependent. We hypothesized that interhospital transfer to endovascular-capable centers would result in treatment delays and worse clinical outcomes compared to direct presentation. Methods -STRATIS was a prospective, multicenter, observational, single-arm study of real-world MT for acute stroke due to anterior-circulation large vessel occlusion performed at 55 sites over 2 years, including 1000 patients with severe stroke and treated within 8 hours. Patients underwent MT with or without IV-tPA, and were admitted to endovascular-capable centers via either interhospital transfer or direct presentation. The primary clinical outcome was functional independence (modified Rankin Score 0-2) at 90 days. We assessed 1) real-world time metrics of stroke care delivery, 2) outcome differences between direct and transfer patients undergoing MT, and 3) the potential impact of local hospital bypass. Results -A total of 984 patients were analyzed. Median onset-to-revascularization time was 202.0 minutes for direct vs. 311.5 minutes for transfer patients (p < 0.001). Clinical outcomes were better in the direct group with 60.0% (299/498) achieving functional independence, compared to 52.2% (213/408) in the transfer group (odds ratio 1.38, 95%CI 1.06-1.79; p=0.02). Likewise, excellent outcome (modified Rankin Score 0-1) was achieved in 47.4% (236/498) of direct patients vs. 38.0% (155/408) of transfer patients (odds ratio 1.47, 95%CI 1.13-1.92; p=0.005). Mortality did not differ between the two groups (15.1% for direct, 13.7% for transfer; p=0.55). IV-tPA did not impact outcomes. Hypothetical bypass modeling for all transferred patients suggested that IV-tPA would be delayed by 12 minutes but MT would be performed 91 minutes sooner if patients were routed directly to endovascular-capable centers. If bypass is limited to a 20-mile radius from onset, then IV-tPA would be delayed by 7 minutes and MT performed 94 minutes earlier. Conclusions -In this large, real-world study, interhospital transfer was associated with significant treatment delays and lower chance of good outcome. Strategies to facilitate more rapid identification of large vessel occlusion and direct routing to endovascular-capable centers for severe stroke patients may improve outcomes. Clinical Trial Registration -URL: http://www.clinicaltrials.gov. Unique identifier: NCT02239640.