Biomechanical investigations are increasingly using commercially available synthetic femurs as surrogates for human cadaveric femurs. However, the rate of force application in testing these artificial femurs appears to be chosen arbitrarily without much consideration to their visco-elastic time-dependent nature. The aim of this study, therefore, was to examine the effect of loading rate on the mechanical behaviour of synthetic femurs. Ten left, medium, fourth-generation composite femurs (Model 3403, Pacific Research Laboratories, Vashon, WA, USA) were fixed distally into cement-filled steel cubic chambers for mounting into a mechanical tester. In randomized order, each of the ten femurs was loaded at rates of 1, 2.5, 5, 7.5, 10, 20, 30, 40, 50, and 60 mm/min to obtain axial, lateral, and torsional stiffness. Axial stiffness showed an aggregate average value of 1742.7 +/- 174.7 N/mm with a high linear correlation with loading rate (R2 = 0.80). Lateral stiffness yielded an aggregate average value of 56.9 +/- 10.2 N/mm and was linearly correlated with loading rate (R2 = 0.85). Torsional stiffness demonstrated an aggregate average value of 176.9 +/- 14.5 N/mm with a strong linear correlation with loading rate (R2 = 0.59). Despite the high correlations between stiffness and speed, practically this resulted in an overall average difference between the lowest and highest stiffness of only 4 per cent. Moreover, no statistical comparisons between loading rates for axial, lateral, or torsional test modes showed differences (p > or = 0.843). Future biomechanical investigators utilizing these synthetic femurs need not be concerned with loading rate effects over the range tested presently. This is the first study in the literature to perform such an assessment.