Abstract. This paper deals with interactive visualization of uncertainty information in probabilistic fiber pathways of the living human brain as measured by diffusion tensor imaging (DTI). We propose a customized direct volume rendering method through which the probabilities of fiber tracts are conveyed by colors and opacities. Furthermore, by rendering multiple scalar-valued data sets simultaneously we are able to visualize sections where multiple fiber tracts intersect and thereby address the methodological issue of crossing fibers. This work is done in close collaboration with DTI domain scientists for hands-on expert knowledge.