This paper presents a new method for integrated time-optimal routing and trajectory optimization of multirotor unmanned aerial vehicles (UAVs). Our approach extends the well-known Traveling Salesman Problem by accounting for the limited maneuverability of the UAVs due to their kinematic properties. To this end, we allow each waypoint to be traversed with a discretized velocity as well as a discretized flight direction and compute time-optimal trajectories to determine the travel time costs for each edge. We refer to this novel optimization problem as the Trajectory-based Traveling Salesman Problem (TBTSP). The results show that compared to a state-of-the-art approach for Traveling Salesman Problems with kinematic restrictions of UAVs, we can decrease mission duration by up to 15\%.