Background and purpose Recent years have seen a rapid increase in the use of resurfacing hip arthroplasty despite the lack of literature on the long-term outcome. In particular, there is little evidence regarding the outcome of revisions of primary resurfacing. The purpose of this analysis was to examine the survivorship of primary resurfacing hip arthroplasties that have been revised. Patients and methods Over 12,000 primary resurfacing hip arthroplasties were recorded by the Australian Orthopaedic Association National Joint Replacement Registry between September 1, 1999 and December 31, 2008. During this time, 397 revisions for reasons other than infection were reported for these primary resurfacings and classified as acetabular, femoral, or both acetabular and femoral revisions. The survivorship of the different types of revisions was estimated using the Kaplan-Meier method and compared using proportional hazard models. Additionally, the outcome of a femoral-only revision was compared to that of primary conventional total hip arthroplasty. Results Acetabular-only revision had a high risk of re-revision compared to femoral-only and both acetabular and femoral revision (5-year cumulative per cent revision of 20%, 7%, and 5% respectively). Femoral-only revision had a risk of re-revision similar to that of revision of both the acetabular and femoral components. Femoral-only revision had over twice the risk of revision of primary conventional total hip arthroplasty. Interpretation Revision of a primary resurfacing arthroplasty is associated with a major risk of re-revision. The best outcome is achieved when either the femoral-only or both the acetabular and femoral components are revised. Technically straightforward femoral-only revisions generally have a worse outcome than a primary conventional total hip arthroplasty.