AbstractDrying fine-grained sediments experience shrinkage and desiccation cracking that may dramatically alter their mechanical and hydraulic properties. This study adopts X-ray computed tomography (CT) to monitor the three-dimensional (3D) internal deformation and strain fields, and their relationships with desiccation crack formation, for drying kaolinite samples contained in plastic containers. Two kaolinite samples, one dried at room temperature and the other oven-dried at 60 °C, were CT scanned at several intervals during the drying process. From sequential CT scans for the same sample, entrained gas bubbles were extracted and used as tracking markers for deformation and strain field measurements. Since the bubble morphology continuously changed during the drying process, an artificial neural network (ANN) model was developed to link bubbles in sequential scans for the same sample. The tracking algorithm was trained with manually linked bubbles and optimised by comparing different combinations of bubble information, e.g. bubble location, size and shape. The drying samples experienced primarily vertical displacement before the air-entry value, while horizontal displacement occurred during vertical crack formation. Internal vertical and horizontal strains were generally uniform, indicating a limited impact of non-uniform sample drying and substrate constraint.