Pain is often the most prevalent symptom among cancer patients referred to hospice or palliative care programs. Tills study was designed to use performance-based testing to evaluate the skills of hospice nurses in assessing the severe pain of a cancer patient and the pain management recommendations they would present to the patient's primary care physician. Twenty-seven hospice nurses (ranging in experience from 1 month to 10 years) were presented with the same standardized patient with cancer pain. In Part A (7 minutes), one of the investigators checked for predetermined behaviors as the nurses performed the clinical pain assessment. In Part B (7 minutes), the nurses answered questions regarding their recommendations for pain management for the patient seen in Part A. In the admission pain assessment, hospice nurses did well in assessing pain intensity (85 %), pain location (70 %), and pain-relieving factors (59%). However, only 48% of the nurses adequately assessed the pain onset, and only 44 % adequately assessed other symptoms the patient might be experiencing. In Part B, 96% of the nurses recommended opioids, 96% recommended the oral route of administration, and 82 % recommended regular dosing of the opioids. Fifty-six percent of nurses included a breakthrough medication in their analgesic recommendations. All of the hospice nurses treated the patient's fear of addiction in an appropriate manner; and 93 % of the nurses recommended increasing the patient's opioid dosage to treat the persisting pain problem. There were no significant differences among nurses with regard to length of time as a hospice nurse or hospice certification on any of the items in either Part A or Part B. Most practicing hospice nurses were judged to be competent in the assessment and management of the severe pain of the standardized cancer patient, although some deficits were noted. Regular oral opioids were the analgesics of choice. Co-analgesics were rarely recommended. (C) U.S. Cancer Pain Relief Committee, 1999.