Fiber-reinforced composites (FRC) have the potential for use as load-bearing orthopaedic implants if the high strength and elastic modulus of FRC implant can be matched with local requirements. This study tested the in vivo performance of novel FRC implants made of unidirectional glass fibers (E-glass fibers in Bis-GMA and TEGDMA polymeric matrix). The implant surface was covered with bioactive glass granules. Control implants were made of surface-roughened titanium. Stress-shielding effects of the implants were predicted by finite element modelling (FEM). Surgical stabilization of bone metastasis in the subtrochanteric region of the femur was simulated in 12 rabbits. An oblong subtrochanteric defect of a standardized size (reducing the torsional strength of the bones approximately by 66%) was created and an intramedullary implant made of titanium or the FRC composite was inserted. The contralateral femur served as the intact control. At 12 weeks of healing, the femurs were harvested and analyzed by radiography, torsional testing, micro-CT imaging and hard tissue histology. The functional recovery was unremarkable in both groups, although the final analysis revealed two healed undisplaced peri-implant fractures in the group of FRC implants. FEM studies demonstrated differences in stress-shielding effects of the titanium and FRC implants, but the expected biological consequences did not become evident during the follow-up time of the animal study. Biomechanical testing of the retrieved femurs showed no significant differences between the groups. The torsional strength of the fixed bones had returned the level of contralateral intact femurs. Both implants showed ongrowth of intramedullary new bone. No adverse tissue reactions were observed. Based on these favorable results, a large-scale EU-project (NewBone, www.hb.se/ih/polymer/newbone) has been launched for development of orthopaedic FRC implants.