Nonunion and segmental bone defects are complex issues in orthopedic trauma. The use of endothelial progenitor cells (EPCs), as part of a cell-based therapy for bone healing is a promising approach. In preclinical studies, culture medium (CM) is commonly used to deliver EPCs to the defect site, which has the potential for immunogenicity in humans. The goal of this study was to find an effective and clinically translatable delivery medium for EPCs. Accordingly, this study compared EPCs delivered in CM, phosphate-buffered saline (PBS), platelet-poor plasma (PPP), and platelet-rich plasma (PRP) in a rat model of femoral critical-size defects. Fischer 344 rats (n = 35) were divided into six groups:​ EPC+CM, EPC+PBS, EPC+PPP, EPC+PRP, PPP alone, and PRP alone. A 5 mm mid-diaphyseal defect was created in the right femur and stabilized with a miniplate. The defect was filled with a gelatin scaffold impregnated with the corresponding treatment. Radiographic, microcomputed tomography and biomechanical analyses were performed. Overall, regardless of the delivery medium, groups that received EPCs had higher radiographic scores and union rates, higher bone volume, and superior biomechanical properties compared to groups treated with PPP or PRP alone. There were no significant differences in any outcomes between EPC subgroups or between PPP and PRP alone. These results suggest that EPCs are effective in treating segmental defects in a rat model of critical-size defects regardless of the delivery medium used. Consequently, PBS could be the optimal medium for delivering EPCs, given its low cost, ease of preparation, accessibility, noninvasiveness, and nonimmunogenic properties.