This paper describes the procedure developed to compare the dynamic behavior of a structure due to an impact load. The comparison procedure is especially useful for controlling the impact level and repeatability during testing, but it can also be used for fine-tuning the process of simplified models or finite element models. Both features provide important engineering value and permit the assessment of transient impact histories, qualitatively and quantitatively. More than 240 soft impacts, from a human impact against glass test campaign, have been used to develop and validate the comparison method.Three indicators have been defined in order to control the most significant impact parameters: the impact level (impulse), duration of the impact (stiffness), and maximum acceleration (peak system response). Based on the knowledge of the impact phenomena, the procedure defines a Similarity Index (SIn) that entails the weighted addition of the indicators. The SIn procedure assigns a numerical value to the difference between two time histories over the duration of the contact, and it is used to compare a series of data from a database of soft human impacts tests versus standard glass curves and curves obtained from a simplified 2DOF model.