The experiments showed the elite women runners to be predominantly midfoot strikers. Their ground reaction forces showed peaks of 3.3 times body weight in the vertical component, 0.8 times body weight in the braking phase, and 0.3 times body weight in the mediolateral direction. The asymmetry in their ground reaction forces was expressed mainly in the mediolateral component. Only minor differences were found between ground reaction forces in racing and training shoes. More abduction during foot placement was associated with greater rearfoot motion and with greater change in mediolateral component of velocity. Among the numerous moderate to high correlations of interest were those between vertical oscillation and peak vertical force, crossover and change in mediolateral velocity, and between movements of the swing leg and mediolateral force values. There were very few correlations between running economy and biomechanical variables, and stride length correlated poorly with stature. The elite runners were shorter in stature, lighter, had shorter legs and considerably less iliac crest fat than a typical non-athletic female population. The runners also had narrower pelves than a student population of similar age. They were reasonably homogeneous and symmetrical in anthropometric dimensions. Compared with a group of elite male runners at the same running velocity, the elite women exhibited more hip flexion, greater angular velocities in hip flexion and extension, and longer stride lengths relative to leg length during running. Their vertical oscillation (expressed relative to leg length) was less than that shown by their male counterparts. On the issue of relative pelvic width, the women had greater relative bitrochanteric and bi-iliac crest widths but similar bispinous widths compared to the elite male runners (all values expressed relative to leg length). Clearly the issue of male-female pelvic width differences at the elite level is not as simple as had previously been suggested.