Background and purpose - Uncemented stems are increasingly used in revision hip arthroplasty, but only a few studies have analyzed the outcomes of uncemented and cemented revision stems in large cohorts of patients. We compared the results of uncemented and cemented revision stems. Patients and methods - 1,668 uncemented and 1,328 cemented revision stems used in first-time revisions due to aseptic loosening between 1999 and 2016 were identified in the Swedish Hip Arthroplasty Register. Kaplan-Meier analysis was used to investigate unadjusted implant survival with re-revision for any reason as the primary outcome. Hazard ratios (HR) for the risk of re-revision were calculated using a Cox regression model adjusted for sex, age, head size, concomitant cup revision, surgical approach at primary and at index revision surgery, and indication for primary total hip arthroplasty. Results - Unadjusted 10-year survival was 85% (95% CI 83-87) for uncemented and 88% (CI 86-90) for cemented revision stems. The adjusted HR for re-revision of uncemented revision stems during the first year after surgery was 1.3 (CI 1.0-1.6), from the second year the HR was 1.1 (CI 0.8-1.4). Uncemented stems were most often re-revised early due to infection and dislocation, whereas cemented stems were mostly re-revised later due to aseptic loosening. Interpretation - Both uncemented and cemented revision stems had satisfactory long-term survival but they differed in their modes of failure. Our conclusions are limited by the fact that femoral bone defect size could not be investigated within the setting of the current study.