Lines of urban public transport with short intervals and high passenger demand have a basic tendency to interval deviations. These are caused by initially slight delays. At following stops, more than the usual number of passengers will have arrived and these will take longer to load. This will lead to a further growth of the delays. Without countermeasures, this process finally leads to the bunching phenomenon. The consequences are operating irregularities and increased waiting times for passengers at the stops. Traffic operators usually tackle this problem by installing slack in the timetables, mostly at the end stops. Thereby delays shall be compensated and the scheduled operations shall be maintained. Due to short intervals in the inner city areas, new approaches refuse the concept of a pre-defined schedule and instead focus on ensuring regular intervals. ^The backward headway method pursues this goal and provides dynamic slack times at one or more control points in the course of the line. At these control points, the vehicles are each delayed by a weighted headway to the following vehicle and thereby continuously converge to an equilibrium value. is a control parameter which is used to wheight the headways. A further parameter specifies a minimum value for the headways at the control point. The method was applied in a slightly modified way in a simulation of the operation of the Viennese tram line 43. Due to high passenger demand and short intervals, this line has strong tendencies to unstable intervals. For the simulation, data on the travel times between individual stops and the number of passengers were needed. Furthermore, data was required on how dwell times depend on the number of persons entering and leaving a vehicle. ^These data were partly collected by measurements, partly provided by the public transport operator Wiener Linien. The travel times between the stops and the dwell times were modeled as random variables and are defined by mean value and standard deviation. The simulation was carried out by means of MATLAB in periods of time which are representative for peak and off-peak hours. The simulation shows that if the backward headway method is applied and and are chosen properly, intervals are considerably more stable, but slightly longer compared to the scheduled current state. The most stable intervals and the shortest possible travel times for the passengers cannot be attained at the same time. Therefore, when choosing the number and position of the control points, a trade-off between this two aims must be found. ^The results of the simulation also show that the backward headway method allows deploying one or two vehicles less, compared to the current operating concept in the rush hour. Stable intervals can be maintained with the reduced number of deployed vehicles. These results are an indication that there is potential for more efficient operations. Finally, another advantage of the method is that manual dispatching could be greatly reduced.