The objectives of this study were to characterize patterns of opioid analgesia in elderly hip fracture patients, to investigate the possible differences in the treatment of cognitively impaired, delirious, or cognitively intact patients, and to study the factors that may affect the doses received by such patients. This retrospective study comprised 184 elderly patients with hip fractures undergoing surgical fixation. Data collection included age, sex, length of stay, type of fracture, cognitive status by mini-mental state examination, assessment of possible delirium by the confusion assessment method, type and doses of opioid received by these patients. We found that the amount of morphine equianalgesic dose differed significantly between demented and non-demented patients (7.5 +/- 1.8 vs. 14.1 +/- 4.9, P<0.001). Patients with cognitive decline or with delirium received only 53 and 34%, respectively, of the amount of opioid that was administered to cognitively intact patients. A significant association was observed between cognitive status, or delirium, and amount of opioid analgesia (P<0.001 and P=0.003, respectively). Other parameters such as age, length of stay and type of fracture, had no effect on the use of opioid analgesia. It is concluded that the management of pain in older persons with hip fracture surgery is suboptimal with regards to insufficient administration of opioid analgesia in demented and delirious patients. The adoption of a standardized protocol for pain control may help in reducing the extent of this problem.