Purpose. To determine whether a mixed course on the Oxford unicompartmental knee arthroplasty (UKA) could improve the non-technical (cognitive) skills of performing UKA. Methods. 120 delegates consisting of consultant orthopaedic surgeons and advanced surgical trainees attended a 2-day course. 104 (87%) of the delegates had performed total knee arthroplasties, whereas 79 (66%) had performed UKAs. The course consisted of didactic lectures, interactive surgical demonstrations, and practical dry-bone skills workshops. Cognitive skills were assessed at the start (day 1) and end (day 2) of the course using 10 multiple-choice questions. The maximum test score was 10. Multilevel modelling was used to compare the pre- and post-course test scores. The pairings of pre- and post-course scores were not known because of a confidentiality clause. Results. Of the 120 delegates, 71 (59%) took the pre-course test and 120 (100%) took the post-course test. The median score improved significantly from 2 (interquartile range [IQR], 0.5–3.5) to 6 (IQR, 4.5– 7.5) [bootstrap p