Objective The identification and treatment of critical illness is often initiated by emergency medical services (EMS) providers. We hypothesized that emergency department (ED) patients with severe sepsis who received EMS care had more rapid recognition and treatment compared to non-EMS patients. Methods This was a prospective observational study of ED patients with severe sepsis treated with early goal-directed therapy (EGDT).We included adults with suspected infection, evidence of systemic inflammation, and either hypotension after a fluid bolus or elevated lactate. Prehospital and ED clinical variables and outcomes data were collected. The primary outcome was time to initiation of antibiotics in the ED. Results There were 311 patients, with 160 (51.4%) transported by EMS. Emergency medical services–transported patients had more organ failure (Sequential Organ Failure Assessment score, 7.0 vs 6.1; P = .02), shorter time to first antibiotics (111 vs 146 minutes, P = .001), and shorter time from triage to EGDT initiation (119 vs 160 minutes, P = .005) compared to non–EMS-transported patients. Among EMS patients, if the EMS provider indicated a written impression of sepsis, there was a shorter time to antibiotics (70 vs 122 minutes, P = .003) and a shorter time to EGDT initiation (69 vs 131 minutes, P = .001) compared to those without an impression of sepsis. Conclusions In this prospective cohort, EMS provided initial care for half of the patients with severe sepsis requiring EGDT. Patients presented by EMS had more organ failure and a shorter time to both antibiotic and EGDT initiation in the ED.