In a prospective study of 313 Dutch cancer patients with chronic pain, the practice of pain treatment was evaluated by means of Donabedian's structure-process-outcome framework. The practice of pain treatment was assessed by: (1) structural resources, describing the setting in which pain treatment is provided; (2) process components, which describe the clinical practice; and (3) outcome measures, which refer to patients' pain intensity, patient satisfaction, or composite pain management index scores. Results showed that 31.4-59.8% of the cancer pain patients received less than optimal pain treatment. Although pain education and refresher courses for health care providers are scarce, structural resources were not the major cause of the suboptimal level of pain treatment. Rather, the major cause was the process components. Only 36.4% of the patients received strong opioids; 23.1% received analgesics "as needed." Patients' pain knowledge was far from optimal (54.8 on a 0-100 scale), and written pain information was given to only 15.8% of the patients. After discharge, only 36.8% of the district nurses were informed about patients' pain. These results emphasize that continuing efforts to improve the practice of pain treatment are needed.