In radiotherapy of gliomas, a precise de nition of the treat- ment volume is problematic, because current imaging modalities reveal only the central part of the tumor with a high cellular density, but fail to detect all regions of microscopic tumor cell spread in the adjacent brain parenchyma. Mathematical models can be used to integrate known growth characteristics of gliomas into the target delineation process. In this paper, we demonstrate the use of di usion tensor imaging (DTI) for simulating anisotropic cell migration in a glioma growth model that is based on the Fisher-Kolmogorov equation. For a clinical application of the model, it is crucial to develop a detailed understanding of its behavior, capabilities, and limitations. For that purpose, we perform a retrospective analysis of glioblastoma patients treated at our institution. We analyze the impact of di usion anisotropy on model-derived target volumes, and interpret the results in the context of the underlying im- ages. It was found that, depending on the location of the tumor relative to major ber tracts, DTI can have signi cant in uence on the shape of the radiotherapy target volume.