Background: Diffusion tensor imaging (DTI) is a relatively new neuroimaging technique that can be used to examine the microstructure of white matter in vivo. A systematic review of DTI studies in schizophrenia was undertaken to test the hypothesis that DTI can detect white matter differences between schizophrenia patients and normal control subjects. Methods: EMBASE, PubMed, Medline, and PsychInfo were searched online and key journals were searched manually for studies comparing anisotropy (a measure of white matter integrity) between patients and control subjects. Nineteen articles were systematically reviewed. Results: Though 16 studies found differences, methodological and data differences prevented a meta-analysis. Fourteen studies found reduced anisotropy in patients; two studies found only a loss of normal asymmetry. The region of investigation varied across studies, however, and when the same region (for example, the cingulum) was examined in different studies, as many failed to find a difference as found one. These inconsistencies may be the result of small sample sizes and differences in methodology. Conclusions: Diffusion tensor imaging has yet to provide consistent findings of white matter abnormalities in schizophrenia. Its potential as a means of examining anatomical connectivity may be realized with the study of larger, more homogenous groups of subjects and with ongoing improvements in image analysis. © 2005 Society of Biological Psychiatry.