Robot cars will likely play an important role in the future. In this paper a visual path following framework for urban environments is experimentally evaluated. The framework's hybrid topological-metric approach for representing the environment provides stable interest points for image-based visual servoing during navigation. The presented experimental results with a robot car show that the framework is robust against changing illumination and moving objects covering up parts of the field of view of the monocular camera. Furthermore, there is no need to perform bundle adjustment nor to use odometry.