Background: A tele-emergency medical service with a remote emergency physician for severe prehospital emergencies may overcome the increasing number of emergency calls and shortage of emergency medical service providers. We analysed whether routine use of a tele-emergency medical service is non-inferior to a conventional physician-based one in the occurrence of intervention-related adverse events. Methods: This open-label, randomised, controlled, parallel-group, non-inferiority trial included all routine severe emergency patients aged ≥ 18 years within the ground-based ambulance service of Aachen, Germany. Patients were randomised in a 1:1 allocation ratio to receive either tele-emergency medical service (n = 1764) or conventional physician-based emergency medical service (n = 1767). The primary outcome was the occurrence of intervention-related adverse events with suspected causality to the group assignment. The trial was registered with ClinicalTrials.gov (NCT02617875) on 30 November 2015 and is reported in accordance with the CONSORT statement for non-inferiority trials. Results: Among 3531 randomised patients, 3220 were included in the primary analysis (mean age, 61.3 years; 53.8% female); 1676 were randomised to the conventional physician-based emergency medical service (control) group and 1544 to the tele-emergency medical service group. A physician was not deemed necessary in 108 of 1676 cases (6.4%) and 893 of 1544 cases (57.8%) in the control and tele-emergency medical service groups, respectively. The primary endpoint occurred only once in the tele-emergency medical service group. The Newcombe hybrid score method confirmed the non-inferiority of the tele-emergency medical service, as the non-inferiority margin of − 0.015 was not covered by the 97.5% confidence interval of − 0.0046 to 0.0025. Conclusions: Among severe emergency cases, tele-emergency medical service was non-inferior to conventional physician-based emergency medical service in terms of the occurrence of adverse events.