BACKGROUND Interactions between the peripheral nervous system and the healing skeleton are poorly understood. Various clinical observations suggest that the nervous system interacts with and promotes fracture-healing. The purpose of this study was to examine the effect of selective sensory denervation on fracture-healing. METHODS Fifty-one Sprague-Dawley rats underwent unilateral placement of an intramedullary rod followed by creation of a standardized femoral fracture. One group of these rats underwent sensory denervation by means of a localized capsaicin injection, and the other did not. Subgroups were allocated for analysis of mRNA expression of collagen I and II and osteocalcin at three, seven, and fourteen days after the fracture. Additionally, histological examination was performed at four weeks; micro-computed tomography, at five weeks; and biomechanical testing, at six weeks. RESULTS The sensory-denervated group had significantly less collagen-I upregulation than the sensory-intact group at three days after the fracture (difference in means, forty-four-fold [95% confidence interval = 22.7 to 65.5-fold]; p < 0.001) and significantly less collagen-II upregulation at seven days after the fracture (difference in means, ninefold [95% confidence interval = 4.3 to 13.8-fold]; p < 0.001). In the sensory-denervated group, the fracture callus had a significantly larger cross-sectional area (difference in means, 15.6 mm(2) [95% confidence interval = 0.78 to 30.5 mm(2)]; p = 0.043) and was less dense. Biomechanical testing revealed that sensory denervation significantly decreased the load to failure (difference in means, 28.7 N [95% confidence interval = 1.2 to 56.2 N]; p = 0.022). CONCLUSIONS Sensory denervation negatively affects fracture-healing. These results offer insight into the nerve-bone interaction following injury.