Diffusion tractography and fMRI data is valuable for pre-surgical planning, but its analysis involves a number of user-dependent decisions. In particular the choice of activation threshold in fMRI maps and the definition of seed region in tractography may impact the results. This work evaluates both the intrinsic variability of probabilistic white matter tract estimation and the inter-user reproducibility of tractography analysis. The former was estimated from repeated identical processing, while for the latter tracts obtained by different users were compared. Achieving a good inter-user reproducibility (up to 85%) is possible, considering that the intrinsic reproducibility ranged between 72% and 89%.