Abstract: Raindrop size distribution (DSD) is a basic characteristic for describing the microphysical process of rainfall. A better understanding of DSD and its variations is not only crucial for improving microphysical parameterization schemes in numerical weather forecasting models, but also important for radar quantitative precipitation estimation. It shows that DSD characteristics are not only related to geographical location, climate, terrain, and humidity, but also vary among different rainfall types and rain rate in the same region. At present, there are still some uncertainties and limitations in the understanding of microphysical characteristics of rainfall in Northeast China, and the microphysical parameterization scheme still lacks accurate description of rainfall microphysical process. Based on observations of the precipitation phenomenon instrument at Xinmin of Liaoning Province in summer, DSD characteristics of different rainfall rate classes are investigated and compared with those of other regions in China. Spectral width of DSD increases with an increase in rain rate (R). The spectral width of raindrops is close to 8 mm when R>20 mm·h^-1. Small drops are predominant in rainfall of Xinmin, but moderate drops make the most significant contribution to total rainfall. Observed DSD samples are also categorized into convective and stratiform rainfall types. The convective rainfall at Xinmin has large raindrop size and low raindrop concentration. Convective rainfall can be identified as continental clusters, with average D_m and lgN_w of 2.14 mm and 3.40, while average D_m and lgN_w of stratiform rainfall at Xinmin are 1.23 mm and 3.30, respectively. The μ-Λ and Z-R relationships for convective and stratiform rainfall at Xinmin are thus fitted. Fitted μ-Λ relationship at Xinmin is similar to that in other regions fitted with data observed by PARSIVEL disdrometers, but different from the empirical relationship fitted from two-dimensional video raindrop spectrometers (2DVD) observations in other regions, and the difference of instruments is the main cause for the discrepancies of μ-Λ relationships. Compared with East China and North China, Xinmin rainfall has larger D_m, lower lgN_w, and higher exponent value of fitted Z-R power-law relationship for convective rainfall, indicating that the radar reflectivity factor at Xinmin increases more rapidly with the increase of rain rate. Using the Z-R empirical formula fitted at Xinmin can reduce the error of radar-based quantitative precipitation estimation. Results would contribute to the understanding of microphysical characteristics of rainfall in Northeast China and the accuracy of radar quantitative precipitation estimation.