Assessments of the bone fracture risk by the standard clinical methods is not sufficiently discriminant. Finite element models (FEM) are accurate when measuring bone strength, but they have not improved bone fracture risk prediction. The risk of fracture is defined as the ratio between external load and the bone strength. Thus, because the FEM estimate the bone strength correctly, a better consideration of the external load should improve the prediction of the risk of fracture. Thus, the aim of this study is to develop an experimental protocol to evaluate bone fragility under realistic loading conditions. The case of a forward fall was considered. Six distal radii were loaded at 75° (between the anterior face of the radius and the ground) with a speed of 2 m/s to simulate a forward fall. Maximum load varied from 852N to 3915N, with an average (±Standard Deviation) of 2134 (±1137). Among the six radii, four fractured and two did not. The current protocol allows the classification of radii in two groups – fractured and non‐fractured – in response to a unique loading case. Having these two groups with known loading conditions (orientation and speed) will be of great interest to assess the predictive capability of FEM.