Context:Existing empirical evidence shows that continuous deep sedation until death is given in about 15% of all deaths in Flanders, Belgium (BE), 8% in The Netherlands (NL), and 17% in the U.K.Objectives:This study compares characteristics of continuous deep sedation to explain these varying frequencies.Methods:In Flanders, BE (2007) and NL (2005), death certificate studies were conducted. Questionnaires about continuous deep sedation and other decisions were sent to the certifying physicians of each death from a stratified sample (Flanders, BE: n = 6927; NL: n = 6860). In the U.K. in 2007–2008, questionnaires were sent to 8857 randomly sampled physicians asking them about the last death attended.Results:The total number of deaths studied was 11,704 of which 1517 involved continuous deep sedation. In Dutch hospitals, continuous deep sedation was significantly less often provided (11%) compared with hospitals in Flanders, BE (20%) and the U.K. (17%). In U.K. home settings, continuous deep sedation was more common (19%) than in Flanders, BE (10%) or NL (8%). In NL in both settings, continuous deep sedation more often involved benzodiazepines and lasted less than 24 hours. Physicians in Flanders combined continuous deep sedation with a decision to provide physician-assisted death more often. Overall, men, younger patients, and patients with malignancies were more likely to receive continuous deep sedation, although this was not always significant within each country.Conclusion:Differences in the prevalence of continuous deep sedation appear to reflect complex legal, cultural, and organizational factors more than differences in patients’ characteristics or clinical profiles. Further in-depth studies should explore whether these differences also reflect differences between countries in the quality of end-of-life care.