Background. The National Confidential Enquiry into Perioperative Deaths recommends that high-dependency (HDU) or intensive care unit (ICU) care is available following arthroplasty. In hospitals without dedicated post-operative care units, patients can be transferred to wards more rapidly, which is associated with adverse surgical outcome, increased morbidity & mortality and unplanned HDU/ICU admission. Pre-operative assessment clinics (POAC) have been demonstrated to reduce these adverse outcomes. We present an evaluation of HDU/ICU admissions and a micro-cost effectiveness evaluation of POAC and planned HDU/ICU admission in hip/knee arthroplasty. Methods. Data were obtained retrospectively for all patients undergoing hip/knee arthroplasty between 01/06/2013–30/06/2014 at North Bristol NHS Trust. n=2258 admissions were linked across coding, ICU (WardWatcher), and Myocardial Ischaemia National Audit Project databases. POAC records and patient notes were hand-searched for n=83 admissions to HDU/ICU. Cost estimates were derived from clinical coding and length-of-stay. The work was performed in STATA and registered under Trust ID 15545. Results. Over eleven months, n=1917 elective arthoplasties were performed, with zero in-hospital deaths and n=68 transferred to HDU/ICU post-operatively. Unplanned HDU/ICU admissions (n=23; 33.8%) were outnumbered by planned admissions (n=38; 55.9%). Hospital length-of-stay was significantly longer (p<0.01) following unplanned HDU/ICU admission, 19.0±22.0 days, compared with planned HDU/ICU admission, 7.5±8.0 days. No significant difference was detected in the proportion of unplanned or planned HDU/ICU admissions that attended anaesthetist POAC (60.9 vs. 68.4%, p=0.59). The total cost of elective arthroplasty with an unplanned (£12200) or planned HDU/ICU admission (£7600) differed by £4500. Conclusions. Our in-hospital mortality compares favourably with published estimates. Unplanned HDU/ICU admission was associated with an increased cost of £4500 per arthroplasty, largely due to increased hospital length-of-stay. 39.1% of patients who required unplanned admission to HDU/ICU were not invited to POAC, which may represent a missed opportunity for reducing post-arthroplasty morbidity and costs. Level of evidence. 2c