BACKGROUND: /st>Major surgery is followed by pronounced sleep disturbances after traditional perioperative care potentially leading to prolonged recovery. The aim was to evaluate the rapid eye movement (REM) sleep duration and sleep architecture before and after fast-track hip and knee replacement with length of stay (LOS) Ten subjects (=60 yr) receiving spinal anaesthesia and multimodal opioid-sparing postoperative analgesia for total hip or knee arthroplasty were included. Ambulatory polysomnography was performed one night before operation at home, continuously during hospitalization, and on the fourth postoperative night at home. Sleep staging was performed according to the American Academy of Sleep Medicine manual. Opioid use, pain, and inflammatory response (C-reactive protein) were also evaluated. RESULTS: /st>The mean LOS was 1.5 (1-2) days. The mean REM sleep time decreased from a mean of 18.2 (9.5-23.5)% of total sleep time to 1.2 (0-5.8)% on the first postoperative night (P=0.002); awake time increased from 19.1 (3.7-44.4)% to 44.3 (12.2-70.6)% (P=0.009); and sleep architecture on the first postoperative night was more disturbed than before operation. Sleep architecture normalized on the fourth postoperative night. There was no association between opioid use, pain scores, and inflammatory response with a disturbed sleep pattern. CONCLUSIONS: /st>Despite ultra-short LOS and provision of spinal anaesthesia with multimodal opioid-sparing analgesia, REM sleep was almost eliminated on the first postoperative night after fast-track orthopaedic surgery but returned to pre-admission levels when at home on the fourth postoperative night.