Cerebral palsy (CP) is associated with movement disorders and reduced muscle size. This latter phenomenon has been observed by computing muscle volumes from conventional MRI, with most studies reporting significantly reduced volumes in leg muscles. This indicates impaired muscle growth, but without knowing muscle fiber orientation, it is not clear whether muscle growth in CP is impaired in the along-fiber direction (indicating shortened muscles and limited range of motion) or the cross-fiber direction (indicating weak muscles and impaired strength). Using Diffusion Tensor Imaging (DTI) we determined muscle fiber orientations and constructed 3D muscle architectures to examine both along-fiber length and cross-sectional area. This is the first approach to our knowledge to use DTI applied to the muscles of subjects with CP. In this project, we present fiber vector data and soleus muscle masks to be used in subsequent analysis and modeling of muscle architecture data in CP.