Introduction. Repeated assessment of pain and other symptoms is required for quality assurance in palliative care. However, physical and cognitive impairment of the patients may impede the use of standardized questionnaires and documentation systems in upalliative care setting. We developed a minimal documentation system (MIDOS) for the specific requirements in this setting. Methods. The German versions of the Brief Pain Inventory (BPI) and the quality of life questionnaire SF-12 were completed for all patients admitted to the palliative care unit. Cognitive impairment was assessed with the Mini Mental State Examination (MMSE). With admission as well as on subsequent consultations patients self-assessed average and maximum pain intensity on numeric rating scales and the intensity of drowsiness, nausea, constipation, dyspnea, weakness, anxiety and well-being on verbal categorical scales. Results. From August 1998 to June 1999 128 patients were documented consecutively. Fifty-nine percent of these patients were treated with WHO-step 3 opioids. Cognitive impairment (MMSE<24) was present in 37% of the patients. Self-assessment with MIDOS was possible for 114 patients at the time of admission, and for 108 patients at the end of therapy. Pain, drowsiness and weakness were documented by most patients, whereas the other symptoms were reported less frequently. Discussion. Factor analysis showed one factor for pain and two factors for the other symptoms. The pain sum score of MIDOS correlated with the factors of the BPI, the symptom sum score of MIDOS correlated with the factors of the BPI and the mental sum score of the SF-12, though on a lower level. MIDOS sum scores showed good pain relief and symptom control for patients discharged home or to other services, whereas the symptom sum score gave an indication of the deterioration in the terminal phase for those patients who died during in-patient treatment. Test-retest stability was good for a subgroup of patients with stable opioid doses. Conclusions. We conclude that MIDOS is a valid instrument for self-assessment of the patient's symptoms and may be used to monitor the efficacy of symptom management.