For a future-oriented statement on the effect of driver assistance systems (DAS) and automated driving functions (AF) on road safety, it is crucial to know and understand today's accident patterns. The accident patterns detected in the process should then be addressed by the systems through their specific functionality. Various methods are available for this purpose not only like simulations but also accident analyses. The latter can cover a broad spectrum, ranging from a general collection of data to “in-depth” analyses. When evaluating the effect of driver assistance systems and automated driving functions by means of accident analyses, which are the focus of this paper, it should be noted that the expected safety effects can vary significantly depending on the accident collective to be analyzed. This applies, for example, when different types of traffic participation (cars, trucks, motorized two-wheelers) are considered individually. It is known from “in-depth” analyses that even this simple distinction reveals special accident patterns for the individual vehicle types, which can often be explained by the different areas in which the vehicles are used. The discussion and analyses regarding assisted or automated driving and their contribution to road safety require a clear understanding of the features and capabilities of the functions. The aim of this paper is therefore, among other things, to point out the need for a clear classification and delimitation of modern driver assistance and automated driving functions as a prerequisite for the analyses of accident data. Based on this classification and methodological assumptions, the analyses clearly show that at the present time the greatest safety benefit can be expected from advanced driver assistance systems. The additional benefit of automated driving functions, especially level 3, in their current design and with a number of influencing variables that cannot yet be quantified, is very limited.