AIM: Today, megaendoprostheses--which were originally designed for osseous defect reconstructions in tumour surgery--are being more frequently used for extensive bone defects in revision arthroplasty. The purpose of this study was to assess the complication rate and the functional results associated with megaendoprosthesis reconstruction of the proximal femur in a non-oncological patient group. METHOD: 28 patients (average age 72; SD 10 years) with a proximal femur replacement were retrospectively (mean follow-up 43 months) evaluated regarding the complication rate. The Harris hip score was used to assess the outcome. The revision surgery was indicated because of large bone defects caused by implant-associated infection (n = 16), periprosthetic fracture (n = 8) or aseptic loosening (n = 4). RESULTS: Overall 8 patients (28.6 %) had to undergo 1 (n = 5) or more (2 n = 2, 4 n = 1) revision surgeries because of dislocation (n = 4), aseptic loosening of the stem (n = 2) and periprosthetic infection (n = 2). A significant pain relief could be achieved from on average 9.0 to 38.7 (according to the Harris hip score). All patients could be mobilised postoperatively, but walking aids were necessary for the majority of patients. CONCLUSION: A proximal femur replacement in revision arthroplasty should be regarded as a salvage procedure for restoration of extremity function. With this procedure it is possible to achieve a--limited--walking ability for patients who were immobilised preoperatively in most cases. Furthermore, pain relief can be achieved. However, the--mostly multimorbid--patients must be informed preoperatively about restrictions in daily life in order to avoid exorbitant expectations.