Background and purpose - Little is known about the role of retained trunnions in revision hip arthroplasties, i.e., when only the femoral head is substituted. Wear (fretting corrosion) and ceramic head fractures are 2 poorly understood concerns related to use, and the role of adapter sleeves has not been defined. In this registry study we assessed the influence of sleeve interposition on re-revision rates in revision hip arthroplasties with retained stems. Confounding factors (demographics, implant-related features) and failures were also analyzed. Patients and methods - We conducted a registry study on 1,127 revised implants (retained trunnion and head exchange). In 26% of implants an adapter sleeve was interposed; in 74% no adapter sleeve was implanted. Demographic and implant-related features were investigated including a descriptive analysis of failures. Results - The mean follow-up of revised implants with and without the use of an adapter sleeve was 3.3 and 5.1 years, respectively. The implant survival without an adapter sleeve was significantly higher, 98.4% (95% CI 96.9-99.8) vs. 95.2% (CI 93.2-96.6) with an adapter sleeve at 5 years. No re-revisions due to adverse local tissue reactions or ceramic head fractures were reported. In order to overcome the different distribution of head materials and head sizes in the two cohorts, only Delta balls were investigated. Interpretation - Adapter sleeve interposition had a minor influence on the revision rates. No adverse local tissue reactions or head fractures occurred.