OBJECTIVE To investigate complications of emergency endotracheal intubation (EEI), possibly facilitated by rapid-sequence induction, in the prehospital critical care setting: 1) the difficulty of intubation; 2) the cardiorespiratory consequences of intubation; 3) the relationship between the occurrence of complications and prognosis. METHODS Prospective non randomized, open study. METHODS All patients treated over a 5-month period by a physician-manned ambulance service and requiring EEI. METHODS Patients were allocated either in with cardiac arrest (CA) group or a group with maintained spontaneous circulation (SC). Difficulty of intubation was assessed by the number of attempts. RESULTS Two hundred and twenty-four consecutive EEI were carried out by physicians (46%) and residents (38%) not trained in anaesthesia, anaesthetists (8%), or nurse anaesthetists (7%). Trachea was intubated after a maximum of three attempts in all patients. Success rate at the first attempt was 91%. It was 92% in CA patients (n = 76) and 90% in SC patients (P = 0.59). Anaesthetic induction, with (n = 112) or without (n = 12) succinylcholine, was used to facilitate 84% of intubations in SC patients. Complications occurred in 30 patients (20%). There was no relationship between the latter and hospital mortality, duration of ventilatory support, duration of stay in the intensive care unit. CONCLUSIONS In this study, EEI in SC patients was frequently facilitated by rapid sequence induction and was associated with a high success rate at the first attempt, as in CA patients. Morbidity was low. All physicians involved in emergency airway management should be skilled in this technique.