Genetic variation in bone-regenerative capacity has not been studied in any animal model system. We developed a "drill-hole" model in the tail vertebra of inbred strains of mice that allows us to reproducibly introduce an injury with a defined boundary and quantify the rate of bone healing using the combination of high-resolution Faxitron X-ray imaging and the ChemiImager 4000 Low Light Imaging System. Using this model, we demonstrate that bone-regenerative capacity is a genetically controlled trait with an estimated heritability of 72%, and that it differs significantly among inbred strains of mice. Of the 12 inbred strains tested, Sencar/PtJ was identified as the most suitable model for the study of hard-tissue regeneration. This strain regained 73% of bone loss 30 days after injury, in contrast to the slow healer, CBA/J, which recovered only 25% of the bone loss during the same period. Bone-regenerative capacity was not correlated with soft-tissue-regenerative capacity, suggesting that different sets of genes may regulate soft- and hard-tissue regeneration. It was, however, significantly correlated with total bone mineral density (R = 0.49, p < 0.01), indicating that high bone density is associated not only with prevention of bone fracture, but also with promotion of bone regeneration.