Unicompartmental knee replacement (UKR) is an option for the treatment of isolated medial compartment osteoarthritis. A commonly perceived potential advantage is that revision of a UKR is straightforward. The purpose of this study was to determine the early outcomes and the level of complexity of revisions of Oxford UKRs performed at our hospital. A retrospective review of a prospective database of all phase III Oxford UKRs was undertaken. This identified 89 Oxford UKRs which were revised at our institution between 2002 and 2008. The median time from the primary procedure to revision was 19 months (interquartile range 2–73 months). Nine were revised to another UKR. Eighty were revised to a total knee replacement (TKR). Fifty-three were revised with primary TKR components. Twenty-seven were revised using stems and/or augments. The median overall tibial component thickness (including augments) was 15 mm. Forty-five knees had an overall tibial component thickness greater than 15 mm. A primary Oxford UKR bearing thickness of greater than 6 mm was associated with an increased likelihood of requiring revision components. On the basis of this review, tibial bone defects were commonly encountered when revising UKRs. Reconstruction with either an augment and a stem, or thick polyethylene component was often required. We recommend that the potential complexity of revision for UKR failure should be borne in mind when considering a primary Oxford UKR.