The relevance of drop shape for erosivity was tested for different combinations of drop sizes and fall heights. For all test combinations together the introduction of the observed drop shape in erosivity parameters only produces minor improvements in the relation between erosivity and detachment and transport by splash. This result is attributed to the dominance of oblate shapes in high velocity conditions. Using small fall heights and low fall velocities as in many rainfall simulators and drop tests it is shown that prolate drops produce a splash detachment which is 2 to 3 times higher than that produced by drops with an oblate shape at impact. As drop production in rainulators or for aggregate stability drop tests may result in more or less uncontrolled variations of drop shape, it is concluded that in addition to other test conditions drop shape should be specified. Moreover it is noted that the erosive capability of prolate drops can explain partly the high splash erosion amounts below vegetation.