Purpose <p>Proximal femur megaprostheses (PFMPs) are used to manage large bone defects in both non-oncological indications (NOI) and oncological indications (OI). However, little is known about the comparative risks of reoperation and functional outcomes between these groups. This study aimed to evaluate the cumulative incidences of reoperation and functional results of PFMPs between NOI and OI.</p> Methods <p>This retrospective, monocentric cohort study included 109 implants between 2005 and 2020 (NOI, <i>n</i> = 42; OI, <i>n</i> = 67). Competing risk analysis was used to estimate and compare cumulative incidence of reoperation and complications, with death as a competing event. The Musculoskeletal Tumour Society Score (MSTS) was retrospectively assessed to compare functional outcomes.</p> Results <p>The estimated cumulative incidence of reoperation at ten years did not statistically differ: 33.5% for NOI vs. 32.7% for OI (HR = 0.90, 95% CI (0.42—0.95), <i>p</i> = 0.791). The estimated cumulative incidence of complications at ten years did not statistically differ (HR = 1.50, 95%CI (0.80—2.80), <i>p</i> = 0.204). The MSTS score at ten years was significantly higher in the OI group (78.2% ± 19.5) compared to the NOI group (48.3% ± 10.9) (<i>p</i> = 0.012).</p> Conclusion <p>PFMPs for NOI demonstrate a comparable risk of reoperation to OI, but with lower long-term functional outcomes.</p>