Paediatric bone sarcomas around the knee are often amenable to either endoprosthetic reconstruction or rotationplasty. Cosmesis and durability dramatically distinguish these two options, although patient-reported functional satisfaction has been similar among survivors. However, the impact on oncological and surgical outcomes for these approaches has not been directly compared. We retrospectively reviewed all wide resections for bone sarcoma of the distal femur or proximal tibia that were reconstructed either with an endoprosthesis or by rotationplasty at our institution between June 2004 and December 2014 with a minimum two year follow-up. Pertinent demographic information, surgical and oncological outcomes were reviewed. Survival analysis was performed using the Kaplan-Meier method with statistical significance set at p<0.05. Thirty eight patients with primary sarcomas around the knee underwent wide resection and either endoprosthetic reconstruction (n=19) or rotationplasty (n=19). Groups were comparable in terms of demographic parameters and systemic tumour burden at presentation. We found that selection of endoprosthetic reconstruction versus rotationplasty did not impact overall survival for the entire patient cohort but was significant in subgroup analysis. Two-year overall survival was 86.7% and 85.6% in the endoprosthesis and rotationplasty groups, respectively (p=0.33). When only patients with greater than 90% chemotherapy-induced necrosis were considered, overall survival was significantly better in the rotationplasty versus endoprosthesis groups (100% vs. 72.9% at two years, p=0.013). Similarly, while event-free survival was not affected by reconstruction method (60.2% vs. 73.3% at two years for endoprosthesis vs rotationplasty, p=0.27), there was a trend towards lower local recurrence in rotationplasty patients (p=0.07). When surgical outcomes were considered, a higher complication rate was seen in patients that received an endoprosthesis compared to those who underwent rotationplasty. Including all reasons for re-operation, 78.9% (n=15) of the endoprosthesis patients required a minimum of one additional surgery compared with only 26.3% (n=5) among rotationplasty patients (p=0.003). The most common reasons for re-operation in endoprosthesis patients were wound breakdown/infection (n=6), limb length discrepancy (n=6) and periprosthetic fracture (n=2). Excluding limb length equalisation procedures, the average time to re-operation in this patient population was 5.6 months (range 1 week to 23 months). Similarly, the most common reason for a secondary procedure in rotationplasty patients was wound breakdown/infection, although only two patients experienced this complication. Average time to re-operation in this group was 23.8 months (range 5 to 49 months). Endoprosthetic reconstruction and rotationplasty are both viable limb-salvage options following wide resection of high-grade bony sarcomas located around the knee in the paediatric population. Endoprosthetic reconstruction is associated with a higher complication rate and may negatively impact local recurrence. Study of a larger number of patients is needed to determine whether the reconstructive choice affects survival