An 8‐mm rat segmental defect model was used to evaluate quantitatively the ability of longitudinally oriented poly(L‐lactide‐co‐D,L‐lactide) scaffolds with or without growth factors to promote bone healing. BMP‐2 and TGF‐β3, combined with RGD‐alginate hydrogel, were co‐delivered to femoral defects within the polymer scaffolds at a dose previously shown to synergistically induce ectopic mineralization. A novel modular composite implant design was used to achieve reproducible stable fixation, provide a window for longitudinal in vivo micro‐CT monitoring of 3D bone ingrowth, and allow torsional biomechanical testing of functional integration. Sequential micro‐CT analysis showed that bone ingrowth increased significantly between 4 and 16 weeks for the scaffold‐treated defects with or without growth factors, but no increase with time was observed in empty defect controls. Treatment with scaffold alone improved defect stability at 16 weeks compared to nontreatment, but did not achieve bone union or restoration of mechanical function. Augmentation of scaffolds with BMP‐2 and TGF‐β3 significantly increased bone formation at both 4 and 16 weeks compared to nontreatment, but only produced bone bridging of the defect region in two of six cases. Histological evaluation indicated that bone formed first at the periphery of the scaffolds, followed by more limited mineral deposition within the scaffold interior, suggesting that the cells participating in the initial healing response were primarily derived from periosteum. This study introduces a challenging segmental defect model that facilitates quantitative evaluation of strategies to repair critically sized bone defects. Healing of the defect region was improved by implanting structural polymeric scaffolds infused with growth factors incorporated within RGD‐alginate. However, functional integration of the constructs appeared limited by continued presence of slow‐degrading scaffolds and suboptimal dose or delivery of osteoinductive signals.