Recent advances in diffusion tensor imaging (DTI) have made it possible to reveal white matter anatomy and to detect neurological abnormalities in children. However, the clinical use of this technique is hampered by the lack of a normal standard of reference. The goal of this study was to initiate the establishment of a database of DTI images in children, which can be used as a normal standard of reference for diagnosis of pediatric neurological abnormalities. Seven pediatric volunteers and 23 pediatric patients (age range: 0-54 months) referred for clinical MR examinations, but whose brains were shown to be normal, underwent anatomical and DTI acquisitions on a 1.5 T MR scanner. The white matter maturation, as observed on DTI color maps, was described and illustrated. Changes in diffusion fractional anisotropy (FA), average apparent diffusion constant (ADC(ave)), and T2-weighted (T2W) signal intensity were quantified in 12 locations to characterize the anatomical variability of the maturation process. Almost all prominent white matter tracts could be identified from birth, although their anisotropy was often low. The evolution of FA, shape, and size of the white matter tracts comprised generally three phases: rapid changes during the first 12 months; slow modifications during the second year; and relative stability after 24 months. The time courses of FA, ADC(ave), and T2W signal intensity confirmed our visual observations that maturation of the white matter and the normality of its architecture can be assessed with DTI in young children. The database is available online and is expected to foster the use of this promising technique in the diagnosis of pediatric pathologies.