The ability to measure different rates of diffusion along different directions is one of the features that distinguish DTI from other imaging methods. It allows to extract and visualize information on tissue microstructure and microdynamics. However, to correctly determine the full diffusion tensor, the so-called b-matrix has to be calculated by taking into account the non-negligible influences of image gradients and cross-terms between imaging and diffusion gradients. In this work validation of this b-matrix correction was investigated by determining self-diffusion coefficients of several isotropic media on a 1.5 T clinical whole-body scanner. To investigate these influences on the measurements of anisotropic media the same experiments were performed with a carrot.