BACKGROUND: Empirical data about end-of-life decision-making practices are scarce. We aimed to investigate frequency and characteristics of end-of-life decision-making practices in six European countries: Belgium, Denmark, Italy, the Netherlands, Sweden, and Switzerland. METHODS: In all participating countries, deaths reported to death registries were stratified for cause (apart from in Switzerland), and samples were drawn from every stratum. Reporting doctors received a mailed questionnaire about the medical decision-making that had preceded the death of the patient. The data-collection procedure precluded identification of any of the doctors or patients. All deaths arose between June, 2001, and February, 2002. We weighted data to correct for stratification and to make results representative for all deaths: results were presented as weighted percentages. FINDINGS: The questionnaire response rate was 75% for the Netherlands, 67% for Switzerland, 62% for Denmark, 61% for Sweden, 59% for Belgium, and 44% for Italy. Total number of deaths studied was 20480. Death happened suddenly and unexpectedly in about a third of cases in all countries. The proportion of deaths that were preceded by any end-of-life decision ranged between 23% (Italy) and 51% (Switzerland). Administration of drugs with the explicit intention of hastening death varied between countries: about 1% or less in Denmark, Italy, Sweden, and Switzerland, 1.82% in Belgium, and 3.40% in the Netherlands. Large variations were recorded in the extent to which decisions were discussed with patients, relatives, and other caregivers. INTERPRETATION: Medical end-of-life decisions frequently precede dying in all participating countries. Patients and relatives are generally involved in decision-making in countries in which the frequency of making these decisions is high.