Assess the risk of complications during endotracheal intubation (ETI) and their association with the skill level of the intubating physician. Prospective cohort study of 136 patients intubated by the intensive care team during a 5-month period. Standardized data forms were used to collect detailed information on the intubating physicians, supervisors, techniques, medications and complications. Canadian academic intensive care unit. All intubations were successful and there were no deaths during intubation. Non-experts were supervised in 92% of procedures. Expert operators were successful within two attempts in 94%, compared to only 82% of non-experts (P = 0.03), with 13.2% of all intubations requiring > or =3 attempts. Furthermore, 10.3% of intubations required 10 or more minutes. Difficult intubation (3 or more attempts by an expert) occurred in 6.6%. Overall risk of complications was 39%, including: severe hypoxemia (19.1%), severe hypotension (9.6%), esophageal intubation (7.4%) and frank aspiration (5.9%). ICU and hospital mortality were 15.4 and 29.4%, respectively. Compared with non-expert intubating physicians, propensity score-adjusted odds ratios (95% confidence interval) for expert physicians were 0.92 (95% CI: 0.28, 3.05, P = 0.89) for any complication, 0.45 (95% CI: 0.09, 2.20, P = 0.32) for ICU mortality and 0.47 (95% CI: 0.13, 1.70, P = 0.25) for hospital mortality. Two or more attempts at ETI was independently associated with an increased risk of severe complications (OR 3.31, 95% CI: 1.30, 8.40, P = 0.01). These prospective data show a high risk of serious complications, and difficult intubations, that are associated with ETI of the critically ill. Artificial airways and complications.