Terminally ill cancer patients near the end of life may experience intolerable suffering refractory to palliative treatment. Although sedation is considered to be an effective treatment when aggressive efforts fail to provide relief in terminally ill patients, it remains controversial. The aim of this study was to assess the need and effectiveness of sedation in dying patients with intractable symptoms, and the thoughts of relatives regarding sedation. A prospective cohort study was performed on a consecutive sample of dying patients admitted to an acute pain relief and palliative care unit within a cancer center. Indications for sedation, opioid and midazolam doses, level of delirium and sedation, nutrition, hydration, rattle, inability to cough and swallow, pharyngeal aspiration, duration of sedation and survival, and use of anticholinergics or other drugs were recorded. Family members were interviewed. Forty-two of 77 dying patients were sedated, and had a longer survival than those who were not sedated (P=0.003). Prevalent indications for sedation were dyspnea and/or delirium. Twelve patients began with an intermediate sedation, and 38 patients started with definitive sedation. The median sedation duration was 22 hours. Opioid doses did not change during sedation. Agitated delirium significantly decreased with increasing doses of midazolam, whereas the capacity to communicate concomitantly decreased. Interviewed relatives were actively involved in the process of end-of-life care, and the decision to sedate, and the efficacy of sedation, were considered appropriate by almost all relatives. Controlled sedation is successful in dying patients with untreatable symptoms, did not hasten death, and yielded satisfactory results for relatives. This study also points to the importance of palliative care and the experience of professionals skilled in both symptom control and end-of-life care.