In two studies the effectiveness of dynamic and multiple static visualizations was investigated for a highly perceptual learning task, namely locomotion pattern classification. In Study 1a, seventy-five students viewed either dynamic, static-sequential, or static-simultaneous visualizations. For tasks with intermediate difficulty dynamic visualizations led to better recognition of the locomotion patterns than static-sequential visualizations, but not than static-simultaneous visualizations. To test whether the presentation of the static-simultaneous visualizations in rows or their permanent visibility was accountable for this effect, three additional static-simultaneous conditions were investigated in Study 1b. Seventy-five students viewed the static-simultaneous visualizations either presented in columns, in matrices, or in circles. The dynamic condition outperformed all three additionally investigated static-simultaneous conditions in the intermediate tasks. Accordingly, for learning how to classify locomotion patterns dynamic visualizations are better suited than most static presentation formats. Nevertheless, presenting static-simultaneous visualizations appropriately can achieve equal results at least for tasks with intermediate difficulty.