Abstract: Raindrop size distribution (DSD) is a key microphysical characteristic of precipitation, as it effectively describes the variation in raindrop number concentration with raindrop size. In-depth DSD research is therefore crucial for understanding the mechanism of precipitation formation and evolution, enhancing the accuracy of weather radar-based quantitative precipitation estimation, and optimizing microphysical parameterization schemes in numerical models. However, studies on DSD characteristics across different regions of Ningxia remain limited, with existing research primarily focusing on Liupanshan. To address this research gap, an analysis of DSD is conducted in 3 representative regions of Ningxia (northern diversion irrigation region, central arid zone and southern mountainous region) using data from DSG5 laser disdrometer. Data are collected from Yinchuan National Reference Climatological Station, Yanchi National Basic Meteorological Station, and Liupanshan National Reference Climatological Station during June-August from 2022 to 2024. The analysis is conducted from multiple perspectives, including overall DSD characteristics, variations under different rain rates, and distinctions across different precipitation types. Results indicate that with increasing elevation, the DSD width and total raindrop number concentration gradually increased, while the mass-weighted mean diameter (D_m) decreased. Liupanshan Station, located at the highest elevation, exhibites a higher concentration of small raindrops, whereas Yinchuan Station, situated at the lowest elevation, shows a higher concentration of medium-sized raindrops. Across different rain rate levels, Liupanshan Station consistently maintained the highest concentrations of both small and large raindrops. During heavy rainfall events, an increased number of large raindrops breaks up during descent, resulting in higher concentrations of small and medium-sized raindrops at both Yinchuan and Yanchi. When the rainfall rate is below 2 mm·h^-1, D_m decreases and the generalized truncation parameter (lgN_w) increases with elevation; when the rainfall rate exceeded 2 mm·h^-1, D_m at Yanchi and Liupanshan surpasses that at Yinchuan Station, while lgN_w is comparatively lower. With increasing elevation, the sample proportion, rainfall contribution, and average intensity of stratiform rainfall increased, whereas those of convective rainfall decreases. For a given shape parameter (μ), the slope parameter (Λ) of Gamma distribution increases with elevation. During convective rainfall events with rainfall rates exceeding 20 mm·h^-1, the empirical relationship overestimates rainfall rates at 3 stations. During a convective rainfull event on 8 August 2024, the fitted relationship derived in this study shows better consistency with measured rainfall rates and smaller deviations compared to conventional empirical methods, particularly for low-intensity rainfall. These results provide important insights into microphysical characteristics of precipitation in different regions of Ningxia and offer valuable implications for improving the accuracy of weather radar-based quantitative precipitation estimation in this region.