Background: As part of a statewide STEMI regionalization program, North Carolina implemented statewide emergency medical service protocols, which encouraged that STEMI patients should be sent directly to a PCI-capable hospital, potentially bypassing hospitals without PCI capacities. We assessed EMS adherence to this protocol, predictors of bypass, and subsequent association of this with patient treatment times and outcomes. Methods: We linked data from the ACTION-GWTG registry and the EMS Pre-hospital Medical Information System (PreMIS) from 06/2008 to 09/2010. Using a Google map application, we selected EMS transported patients that either 1) bypassed a closer non-PCI hospital and went directly to a PCI center, or 2) were first taken to a non-PCI center and subsequently transferred for PCI.’ We determined predictors of bypass using multivariable logistic regression modeling accounting for clustering within hospital referral regions. Time from first-medical-contact (FMC) to PCI/reperfusion and in-hospital mortality were compared between groups. Results: Among 1224 eligible STEMI patients, 765 (63%) underwent bypass to a PCI facility while 479 (37%) were first treated at a non-PCI hospital and then transferred for PCI. Adjusted predictors of undergoing bypass were white race (OR 1.37, 95% CI 1.02-1.84), chief complaint of chest pain (OR 2.08, 95%CI 1.46-2.95), having received a pre-hospital 12-lead electrocardiogram (OR 2.14, 95%CI 1.10-4.15), cardiogenic shock (OR 1.82, 95% CI 1.22-2.72), and a prior history of PCI (OR 1.66, 95% 1.11-2.50). Time from FMC to PCI was 95 min (IQR 77-117) in the bypass group vs 179 min (IQR 138-288) in the non-bypass group, p-value for difference <0.0001. Time from FMC to initial reperfusion therapy (PCI or fibrinolysis) was 94 min (IQR, 76-116) in the bypass group vs. 124 (IQR, 67-179) in the non-bypass group, p for difference <0.0001. Crude inhospital mortality was lower in the bypass group vs. the non-bypass group (6.3% vs. 9.4%, p-value for difference = 0.046). Conclusions: After implementation of a statewide STEMI regionalization program, 2/3 of EMS transported patients bypassed a closer non-PCI center and went directly to a PCI-capable center. Patients who bypassed and went directly to PCI capable centers had significantly more rapid reperfusion times and lower mortality compared with patients not undergoing EMS.