Differences in the general focus of care among hospitals, nursing homes, and homes may affect the adequacy of end-of-life decision making for the dying. We studied end-of-life decision-making practices for cancer patients who died in each of these settings and assessed the impact of the Liverpool Care Pathway for the Dying Patient (LCP), a template for care in the dying phase. Physicians and relatives of 311 deceased cancer patients completed questionnaires. The LCP was introduced halfway through the study period. During the last three months of life, patients who died in hospital received anticancer therapy and medication to relieve symptoms more often than those in both other settings. During the last three days of life, patients who died in the hospital or nursing home received more medication than those who died at home. The LCP reduced the extent to which physicians used medication that might have hastened death. Relatives of patients who died in the hospital tended to be least positive about the patient's and their own participation in the decision making. We conclude that cancer patients who die in the hospital are more intensively treated during the last phase of life than those who die elsewhere. The LCP has an impact on the use of potentially life-shortening medication during the dying phase. Communication about medical decision making tends to be better in the nursing home and at home.