Abstract Introduction While re-revision total knee arthroplasty (ReRTKA) steadily increases, the ease and bone-sparing removal of RTKA implants is gaining more and more in importance. Biomechanical data suggest that cemented conical stems can be removed significantly easier than cylindrical stems. However, no clinical evidence exists supporting this observation. Aim of this study was to compare the revisability and need for osteotomy (OT) between removals of well-fixed cemented conical vs. cylindrical RTKA stems. Materials and methods 55 removals of well-fixed full-cemented RTKA stems (29 knees) performed between 2016 and 2018 were retrospectively analyzed. Main outcome variables were: bone loss, fractures, osteotomy incidence, surgery duration, early postoperative complications (EPC), hemoglobin drop and blood transfusion. SPSS was used for the statistical analysis. Results 44.8% were conical, 48.3% cylindrical, and 6.9% combined stem designs. Causes for re-revision were PJI (75.9%), malposition (17.2%) and persistent pain (6.9%). 10 stem removals (18.2%) required an OT (four femoral, six tibial): eight stems (14.5%) had cylindrical and two (3.6%) conical designs (P = 0.041). Fractures were noted solely in removals without OT (11.1% vs. 0%,). There was a tendency to more bone loss in cylindrical stem revisions (53.8% vs. 32%, P = 0.24). A longer overall surgery time was observed in revisions of cylindrical stems (+ 37 min, P = 0.05). There was higher hemoglobin drop and need for blood transfusion in revisions of cylindrical stems or after OT but without reaching statistical significance. The EPC rates were slightly higher in ReRTKA on cylindrical stems (P = 0.28). Conclusion Well-fixed cemented conical stems may be revision friendlier with less demands on OT and shorter overall surgery time than cemented cylindrical stems.