Background In regional systems of ST-segment elevation myocardial infarction (STEMI) care, patients presenting to hospitals without percutaneous coronary intervention (PCI) are transferred to PCI-capable hospitals for primary PCI. Repatriation, a practice whereby such patients are transferred back to non-PCI referral hospitals after reperfusion is prevalent in many jurisdictions, yet little is known of this practice and its safety. Methods We studied 979 consecutive STEMI patients transported from the emergency department and catchment area of two non-PCI hospitals in Ontario, Canada to a regional PCI-hospital for primary PCI between January 2008 and June 2014. Logistic regression modeling was performed to determine factors associated with delayed repatriation beyond 24 hours and to evaluate the association between repatriation and index-admission mortality. Results Eight hundred and fifteen (83.2%) patients were repatriated with 524 (65.2%) patients repatriated within 24 hours. Factors independently associated with delayed repatriation included systolic blood pressure (OR 1.03 per 5 mmHg decrease, 95% CI 1.01-1.06, P=.04), requirement for mechanical ventilation (OR 24.9, 95% CI 5.4-115.3, P<.0001), ventricular arrhythmia (OR 3.0, 95% CI 1.3-6.6, P=.01), infarct-related artery (P=.03), final TIMI flow grade (P=.01) and access-site complications (OR 2.36, 95% CI 1.04-5.4, P=.04). After repatriation, 9 (1.3%) patients returned to the PCI-hospital for urgent care, and 16 (2.0%) died during index-admission. After adjustment, repatriation was not associated with increase in index-admission mortality (adjusted OR 0.46, 95% CI 0.16-1.32, P=.15). Conclusions In a regional STEMI care system in Ontario, Canada, patients are routinely repatriated to non-PCI hospitals after primary PCI. This practice was associated with very low and acceptable rate of return to the PCI-hospital during index-admission without an adverse impact on short-term outcomes.