The induced membrane technique is now widely used for pediatric diaphyseal bone loss due to various etiologies. Although consolidation rates remain satisfactory, complications, and healing delays may occur requiring additional procedures. We studied a series of induced membrane bone reconstructions in which the second stage included an embedded endomembranous non vascularized fibular shaft, in addition to iliac bone grafts. The purpose of this study was to analyze the results in terms of bone consolidation and complications. This is a retrospective comparative and multicentric study of 32 children with large bone loss treated with the induced membrane reconstruction technique. Patients were divided into 2 groups according to the graft used during the second stage. The first group (G1) of 16 patients had a nonvascularized fibula embedded inside the membrane in addition with the corticocancellous grafts from the iliac crest. The second group (G2) of 16 patients underwent reconstruction using the original technique, with iliac crest graft only. The 2 groups were similar in terms of etiologies of bone loss and follow-up (mean: 44mo for G1 and 49mo for G2). Mean bone losses were 15.4cm (range: 2 to 25; SD: 5.6) for G1 and 10.6cm (range: 3 to 19; SD: 5.2) for G2. In the first group, all patients healed primarily, with a mean time of 5.9 months (range: 4 to 8; SD: 1.6). In the second group, 2 of 16 patients did not healed; for the others 14, healing mean time was 6.9 months (range: 3 to 12; SD: 2.7). The short-term and long-term complications rates were 38% to 19% for G1 and 50% to 31% for G2, respectively. Regarding the donor site, the fibulas reconstructed spontaneously with a mean time of 4.8 months (range: 3 to 6; SD: 1.2). The integration of a nonvascularized fibula during the second stage of the induced membrane technique appears to improve the consolidation rate in the pediatric population. Level III-Retrospective comparative study.