Measurement of the velocity of propagation and attenuation of ultrasound (200 kHz) is believed to be a useful non-invasive technique for assessing the mechanical properties of bone. A new method for the determination of ultrasound velocity and attenuation of longitudinal waves in cortical bone was used in vivo and in situ on intact and fractured human tibiae. The measured ultrasound attenuation and velocity were found to be unaffected by the soft tissue between transducers and bone. The ultrasound velocity in vivo on control tibiae was and the attenuation was . The ultrasound velocity in fractured tibiae was considerably lower 1 week after fracture (), but had significantly increased after 3 weeks (to ). A higher attenuation was measured 1 week after fracture (), but it had decreased again 3 weeks after fracture (). In situ studies under well-defined conditions confirmed the in vivo results. The effects of internal plate fixation and gradually cutting through the cortex on the ultrasound velocity and attenuation were studied in situ. These results demonstrate the clinical potential of this technique for the non-invasive assessment of bone fracture healing.