Most running shoe investigations have used the same standard procedure for the evaluation of the shoes: the runners are filmed from behind and a film analysis is carried out digitizing markers at the heel counter of the shoe and on the lower leg. The angular displacement of these markers relative to the horizontal or the vertical is assumed to be an indicator for various sports injuries. The goal of this investigation was to measure the movement of the heel counter as well as the movement of the heel inside the shoe. First, the influence of the size of different heel counter windows was controlled and found negligible for the test conditions of this study. Second, 15 subjects performed the following procedure: running (a) barefoot, (b) with shoes with windows, and (c) without windows. Overall, the heel was found to move similarly but not identically to the heel counter. The maximum change of pronation was (a) 13.7 +/- 3.7 degrees, barefoot; (b) 14.1 +/- 3.8 degrees for the shoe with windows and 12.1 +/- 3.7 degrees for the heel inside these shoes; and 14.9 +/- 4.2 degrees for the shoes with no windows. To achieve a general impression of a shoe in the sense of a qualitative description, the previous method without heel counter windows still seems adequate. However, for a detailed analysis of quantitative nature, it is important to use the method with heel counter windows.