Diffusion tensor imaging (DTI) has been widely used to study rodent brain development. Nevertheless, the parameters derived from DTI are sensitive to, but non-specific to, the tissue’s microstructure. Recently, NODDI (neurite orientation dispersion and density imaging) has been proposed. We aimed to estimate the real input of NODDI derived parameters in rodent brain development. ODI appears more accurate and specific to reflect GM (increase with dendritic arborization) and WM (decrease with myelination) development than FA and could be a very important parameter in the assessment of perinatal brain injuries. Conclusion about the other NODDI estimates requires further experiments.