Quantifying the changes in morphology of wheat grain across development is essential to better understand the factors limiting its growth. In this work, we investigated such changes using X-ray micro computed tomography (µCT) and 3D digital image processing. We developed a robust semi-automatic image processing methods for identification of 3D wheat grain and extraction of grain morphometry. We also proposed an approach to construct an average grain as a representative shape for each developmental stage. The quantitative description of morphology and of the average grain enlights the dierences of growth within the grain. These first results will help build a developmental atlas of wheat grain based on 3D imaging data.