Fracture healing has traditionally been monitored subjectively using manual manipulation of the fracture site and evaluation of radiographic images. A more objective method of monitoring would provide obvious advantages, allowing healing progress to be quantitatively assessed and so providing the opportunity for early detection of problems. A tibia phantom was used to investigate whether the longitudinal propagation velocity of ultrasound across a fracture site could be used quantitatively to assess fracture healing. The characteristic of fracture healing simulated by the phantom was the changing gap between the bone ends at the fracture site. The ultrasound velocity was measured using a recently developed machine, the SoundScan 2000 (Myriad Ultrasound Systems Ltd, Israel). The precision of the SoundScan 2000 was found to be 0.4% in vitro. Ultrasound velocity predicted the simulated fracture gap with a high degree of accuracy (R2 = 0.994). The measured and the theoretically calculated velocity for different widths between the simulated bone ends was found to be highly correlated with a coefficient of determination of 0.998. This result shows that the use of quantitative ultrasound to monitor fracture healing warrants further investigation in vivo.