The purposes of this study were to examine the relationship between: (1) nurses' ratings of pain and corresponding administration of pain medication to elderly long-term care residents, and (2) cognitive status of the elderly and pain medication orders/administration. Participants were 83 residents, 60 years of age and older, in two groups: cognitively impaired (n = 64), and cognitively intact (n = 19). For comparison purposes, 19 of the cognitively impaired subjects were matched on age and diagnosis to provide control for potentially painful conditions. A retrospective medication review of the resident's charts was conducted to compare medication orders and administration on analgesics that were scheduled and p.r.n. (given as needed). The pain ratings of 25 RNs using a visual analogue scale were correlated with pain medications given to the resident on the day of the rating. Results indicated that RNs' ratings of resident pain and the administration of pain medications were not significantly correlated. In addition, cognitively impaired residents were prescribed significantly less scheduled medication and received significantly less pain medication (either p.r.n. or scheduled) than the cognitively intact elderly. Implications for practice and research are discussed.