Aims: To assess the results of the Oxford unicompartmental knee prosthesis, and compare these with other prostheses. Methods: Ninety-seven prostheses (87 medial; 10 lateral) in 86 patients were evaluated with the HSS (Hospital for Special Surgery) score after 2–14 years (mean follow-up: 6 yrs-9 mos). Results: Five prostheses were lost to follow-up. Eight patients died after a mean of 7 years, none of them had undergone a revision. Fourteen revisions (of which 1 bilateral unicompartmental knee prosthesis), 11 medial and 3 lateral, were performed. The mean HSS-score of the 69 UKPñs is 178.8 (80% excellent, 10% good, 4% fair, 6% poor). Conclusions: With proper patient selection (i.e. degenerative medial unicompartmental arthritis, good range of motion, sedentary occupation and with a light body weight) and a consistent operating technique, the results are good. It is the þrst-choice prosthesis for the relatively young patient (with regard to later revisions). Since we also obtain good results in the elderly, it seems a good choice of treatment in this group.