In recent years, the detection frequency of microplastics (MPs) in groundwater systems has been steadily increasing, drawing widespread attention to their potential hazards. However, current understanding of the migration and transformation mechanisms of microplastics in groundwater within semi-arid regions, as well as the associated risks, remains limited. This study systematically investigated the abundance, polymer types, and size distribution of microplastics in groundwater within the Bojianghaizi Basin of Ordos City, China. Employing hydrogeochemical methods and ecological risk modeling, it analyzed spatial variation patterns, migration and transformation mechanisms, and ecological risks associated with these microplastics. Results indicate that wellhead management practices significantly influence microplastic abundance in groundwater. Sealed wells averaged 5.09 particles/L, with polypropylene (PP) and polyvinyl chloride (PVC) as the predominant polymer types, accounting for 33.13% and 18.66% respectively. Microplastics predominantly existed as fragments (99.58%), with particle sizes primarily concentrated between 20-100 μm (90.94%).) Open wells exhibited abnormally elevated microplastic abundance (up to 141.99 particles/L) due to atmospheric deposition input. Groundwater runoff had no significant effect on the enrichment of MPs ranging from 20 to 500 μm. The large pore size of the vadose zone facilitates the downward migration of surface-derived microplastics into groundwater, while high redox potential (Eh) values in groundwater promote their oxidation, decomposition, and fragmentation. The PCI index (Pollution Composite Index) indicates that although the absolute abundance is low, the disproportionately high proportion of highly toxic polymers places microplastics in the groundwater of the Bojianghaizi watershed at a relatively high ecological risk level overall. This study provides a scientific basis for understanding the mechanisms and assessing the risks of microplastic pollution in groundwater systems in semi-arid regions.