Inner city intersections are a challenging scenario for human drivers as well as for the development of autonomous vehicles. This is especially the case for unsignalized intersections where the right before left rule applies. At these intersections, ambiguous situations can arise. In this chapter, we cover two aspects of this intersection type: First, we use driving data from a field study conducted in inner city traffic to analyze the relationship between intersections and human driving behavior. For that, we describe the intersection, its surrounding environment and the traffic there by features that constitute an intersection’s complexity (e.g. street width, visibility conditions, number of cooperation vehicles). With those we are able to predict features describing the driving behavior reliably. Second, we propose a decision making algorithm for unsignalized inner city T-junctions. The algorithm is modeled as a discrete event system and does not rely on any explicit communication. Instead, only the observable state is used. This includes the map, the positions and velocities of the cooperation vehicles and the driving pattern. We introduce the algorithm in detail and present results of a comprehensive simulation for validation. The algorithm is able to drive through all situations in the simulation safely.