The Jinsha River serves as a vital ecological barrier in the upper reaches of the Yangtze River. This area has been abundant in aquatic biological resources but is being severely impacted by the cascade hydropower stations along its watercourse. eDNA metabarcoding can efficiently and standardly obtain biodiversity data, widely recognized by both domestic and international researchers as a technology capable of replacing traditional methods for biodiversity monitoring. To explore the spatial pattern of benthic macroinvertebrate diversity and the related influencing factors in the Jinsha River mainstream, eDNA metabarcoding and traditional methods were used to sample the macroinvertebrates at 16 sampling sites along the mainstream of the middle and lower reaches of the Jinsha River in February 2021. Water physicochemical variables were simultaneously measured, and environmental data of hydropower stations were collected. The eDNA metabarcoding technique detected a total of 46 operational taxonomic units (OTUs) across 4 phyla, 6 classes, 15 orders, and 24 families of benthos, while traditional morphological survey methods identified 29 genera within 3 phyla, 5 classes, 9 orders, and 19 families. Both methods found that Diptera had the highest diversity. The middle and lower reaches of the Jinsha River showed significant differences in macroinvertebrate community composition, as revealed by both eDNA sequences and morphological abundance data. In addition, eDNA methods further detected the remarkable differences in the dominant taxa and alpha diversity between middle and lower reaches. Canonical correspondence analysis identified altitude, permanganate index, and the average annual power generation of hydropower stations in the upper reach of sampling sites as the key environmental factors affecting macroinvertebrate communities. Our results highlighted the significant differences in macroinvertebrate diversity between the middle and lower reaches of the Jinsha River and cascade hydropower development is the major stressor on macroinvertebrates. Our findings provide a reference for the improvement of biodiversity conservation and monitoring methods for aquatic organisms in the Jinsha River.