Bone, with a hierarchical structure that spans from the nano-scale to the macro-scale and a composite design composed of nano-sized mineral crystals embedded in an organic matrix, has been shown to have several toughening mechanisms that increases its toughness. These mechanisms can stop, slow, or deflect crack propagation and cause bone to have a moderate amount of apparent plastic deformation before fracture. In addition, bone contains a high volumetric percentage of organics and water that makes it behave nonlinearly before fracture. Many researchers used strength or critical stress intensity factor (fracture toughness) to characterize the mechanical property of bone. However, these parameters do not account for the energy spent in plastic deformation before bone fracture. To accurately describe the mechanical characteristics of bone, we applied elastic–plastic fracture mechanics to study bone's fracture toughness. The J integral, a parameter that estimates both the energies consumed in the elastic and plastic deformations, was used to quantify the total energy spent before bone fracture. Twenty cortical bone specimens were cut from the mid-diaphysis of bovine femurs. Ten of them were prepared to undergo transverse fracture and the other 10 were prepared to undergo longitudinal fracture. The specimens were prepared following the apparatus suggested in ASTM E1820 and tested in distilled water at 37°C. The average J integral of the transverse-fractured specimens was found to be 6.6 kPa m, which is 187% greater than that of longitudinal-fractured specimens (2.3 kPa m). The energy spent in the plastic deformation of the longitudinal-fractured and transverse-fractured bovine specimens was found to be 3.6–4.1 times the energy spent in the elastic deformation. This study shows that the toughness of bone estimated using the J integral is much greater than the toughness measured using the critical stress intensity factor. We suggest that the J integral method is a better technique in estimating the toughness of bone.