The growth trajectory of hailstones in clouds determines the ground intensity and spatial distribution of hailfall. A systematic study of hail trajectories can help improve the current scientific understanding of the mechanisms by which hail forms in semi-arid regions of China and, in doing so, improve the quality of hail forecasts and warnings and help to prevent and mitigate disasters. In this study, the WRFv3.7.1 model was employed to provide the background field to drive the hailstone trajectory model. Cluster analysis was then used to classify hail trajectories to investigate the characteristics of different types of hail trajectories and the microphysical characteristics of hail formation. The differences in hail trajectories might be mainly due to differences in the background flow fields and microphysical fields of hail clouds in different regions. Comparative analysis revealed that as the maximum particle size of ground hailfall increased, the maximum supercooled cloud water content and the maximum updraft velocity for the formation and growth of hailstone increased. The larger the size when the hailstone reaches its maximum height, the larger the ground hailstone formed. Overall, the formation and growth of hailstone are caused by the joint action of the dynamical flow field and cloud microphysical processes. The physical processes of hailstone growth and main growth regions differ for different types of hail trajectories. Therefore, different catalytic schemes should be adopted in artificial hail prevention operations for different hail clouds and trajectories due to differences in hail formation processes and ground hailfall characteristics.