Plant diversity in riparian areas is an important topic in restoration ecology and biodiversity conservation science. The Aqikesu River in Xinjiang Uygur Autonomous Region, China, a typical inland river in this arid desert region, is the only river located in the Ebinur Lake Wetland National Nature Reserve and thus plays a crucial role in maintaining balanced and sustainable development of the local ecosystem. Exploration of plant diversity patterns and influencing mechanisms in the riparian zone in this desert area provides scientific baseline data related to the conservation of riparian biodiversity and scientific management of this reserve. However, the patterns of biodiversity and influencing mechanisms in the riparian zone along this river remain unclear. This study used four transects established perpendicular to the Aqikesu River in north and south of Dongdaqiao Station. The four transects were located 0.1-0.2 km, 0.2-0.4 km, 0.9-1.5 km, and 2-3 km from the river and named T_0.2, T_0.4, T_1.5, and T_3.0, respectively. Plots were established in each transect based on the appearance and characteristics of the plant communities, that is 10 m × 10 m for trees, 5 m × 5 m for shrubs, and 1 m × 1 m for herbs. Specifically, T_0.2 included nine plots (three, four and two for trees, shrubs and herbs, respectively) and T_0.4 had six plots (two plots each for trees, shrubs and herbs). T_1.5 had eight plots (two plots each for trees and herbs, and four for shrubs). T_3.0 had eight plots (two plots for trees and three plots each for shrubs and herbs). Geographic data were recorded in each plot; abundance, height, and diameter at breast height of trees were surveyed and recorded along with the abundance of shrubs and herbs. Soil samples were collected in each plot and soil water content (SWC), total salt (TS), pH, soil organic matter (SOM), and soil phosphorus (P) were measured in the lab. First, based on field and experimental data, the intrinsic and α diversity indices of T_1.5 and T_3.0 transects were all higher than those of T_0.2 and T_0.4 transects. Secondly, T_0.4 transects exhibited the lowest diversity; the Cody and Wilson index of T_0.4 was significantly lower than in other transects (P <0.05), and the Jaccard and Sørenson indices of T_0.4 were significantly higher than that of other transects (P < 0.05). Compared with T_0.2, the T_0.4 transects had greatly reduced herbaceous flora, SWC, TS, and soil organic matter, which indicated the transitional nature of the community. Therefore, the T_0.4 transect could function as a reference for determining the width of the riparian zone. Third, as distance from the river increases as one proceeds from T_0.2 to T_3.0, the effects and influence of soil conditions on plant diversity gradually become simpler. In the nearest distance away from the river (T_0.2), plant diversity is constrained by poor nutrition (low carbon/ phosphorus ratio (C/P) and low SOM), high soil moisture, and high salt content. Farther from the river in the T_0.4 and T_3.0 area, only soil P and SWC were observed to gradually and prominently affect plant diversity more and more as drought stress increases with distance from the river. Fourth, based on plant diversity traits of T_1.5 and T_3.0, the highest plant diversity occurs when soil conditions in the study area have a an SWC between 7.0% and 7.5%, soil C/P between 26.1 and 30.2, and soil salt content of less than 1.0%. Finally, specific suggestions for determining the appropriate width of riparian and buffer zones and suggestions for riparian zone management, vegetation resource protection, and ecosystem restoration were proposed to provide a theoretical and scientific basis for effective biodiversity maintenance, resource management, and habitat damage mitigation in riparian zones.