As demographic trends show, the demand for home health care services will rise in future. Currently, the routing of the nurses is performed manually by the main service providers in Austria. This leads to a time-consuming process with a presumably suboptimal outcome. We present a model formulation and a metaheuristic solution approach, based on Variable Neighborhood Search, for optimizing the daily scheduling of the nurses. The objective function minimizes the traveling time of the nurses and the dissatisfaction level of clients and nurses. A feasible solution has to observe working time regulations, hard time windows, mandatory breaks, and a feasible assignment of nurses to clients. The proposed method finds the global optimal solutions for small problem instances. In extensive numerical studies it is shown that the algorithm is capable to solve real life instances with up to 512 home visits and 75 nurses. A comparison with a real life routing plan shows large savings potentials.