Delirium is a major adverse postoperative event in elderly patients. Incidence rates of postoperative delirium are difficult to determine. Because of the accuracy, brevity, and ease of use by clinical interviewers, the Confusion Assessment Method (CAM) has become widely used. This study used the CAM to determine the rate of postoperative delirium in patients undergoing total hip arthroplasty (THA) and total knee arthroplasty (TKA) procedures under regional anesthesia. Following Institutional Review Board approval, a prospective study of 20 patients per group ages 70 and above undergoing unilateral THA or TKA was initiated. Both groups received a combined spinal–epidural, postoperative patient-controlled epidural anesthesia, and postoperative oral opioids. Patient interviews occurred five times: once preoperatively and two times each on postoperative days 1 and 2. Only two patients were assessed as delirious according to the CAM method (one in each group; 5%). Patient assessment by other clinicians indicated that five additional patients experienced acute change in mental status; however, these patients were not delirious at the times of the study interviews. The rate of delirium in the elderly after arthroplasty performed under regional anesthesia is very low. Reasons for this include patient selection criteria and anesthesia type. The study excluded patients with several proposed risk factors for postoperative delirium: prior history of dementia, history of mental illness, and use of benzodiazepines. The use of regional anesthesia may have also reduced the occurrence of postoperative delirium.