STUDY OBJECTIVE To determine whether use of intraoperative neuraxial anesthesia would decrease the frequency of ICU admission postoperatively in orthopedic patients. DESIGN Retrospective chart review. SETTING University hospital. MEASUREMENTS We reviewed the data from all patients who underwent elective total knee or hip replacements from January 1, 1999 to September 30, 2000 and were preoperatively scheduled for ICU admission following surgery. We recorded patient demographic data, presence of comorbidities, intraoperative data, and location of postoperative recovery. MAIN RESULTS Of the 361 patients having total knee and hip replacements, 88 patients were scheduled for postoperative ICU admission. Forty-five patients underwent neuraxial (epidural or spinal) anesthesia, 38 patients received general anesthesia, and 5 patients received a combination of both neuraxial and general anesthesia. Patients who received neuraxial anesthesia had a significantly lower rate of actual ICU admission when compared with those who received intraoperative general anesthesia (11/45 or 24% vs. 22/38 or 58%, p = 0.002). Patients who received neuraxial anesthesia had shorter anesthesia and surgical times when compared with those who received general anesthesia. CONCLUSIONS Intraoperative neuraxial anesthesia in higher-risk patients undergoing elective hip or knee replacement surgery is associated with a decrease in anticipated ICU admission postoperatively. A causal relationship cannot be determined with this type of study and further research is needed to better understand this association.