Several aspects of dynamic foot function were investigated using ground reaction forces obtained from five runners performing in five different shoes. A minimum of eight trials were necessary in order to obtain stable subject-condition values. Statistically significant subject-shoe interactions occurred for all parameters. It was therefore concluded that a knowledge of shoe characteristics independent of subject characteristics provided little useful information regarding the effects of the shoes tested on selected foot mechanics of the runners. The various subject-conditions were evaluated in an attempt to determine which was the best shoe. Selected descriptive data are presented and the 'best' shoe concept discussed.