Abstract: The performance of MRR (Micro Rain Radar) in estimating DSD (raindrop size distribution) and integral rainfall parameters (rainfall intensity R, liquid water content W, radar reflectivity factor Z, mean mass-weighted raindrop diameter D_m, and intercept parameter N_w of the normalized gamma distribution) was comprehensively evaluated using coincident data obtained from a 2DVD (two-dimensional video disdrometer) and an XPOL (X-band dual-polarization radar) in Beijing in the warm season (May–September) during 2016–2018. A comparative analysis between standard MRR products (hereinafter referred to as AVG data) and reprocessed data (hereinafter referred to as REP data) derived from raw MRR spectra was conducted to assess the accuracy of the MRR in estimating stratiform and convective rain types. The key findings were as follows: (1) The retrieval accuracy of the DSD and rainfall parameters for the REP data outperformed that from the AVG data, with device performance being superior for stratiform rain than for convective rain. The Z bias improved from −23.74 dB (for AVG data) to 0.74 dB (for REP data) in the case of convective rain. (2) The AVG data systematically underestimated the DSD and rainfall parameters compared to 2DVD measurements, an underestimation that intensified as rainfall intensity increased. In contrast, the REP data exhibited not only higher correlation coefficients (Z: 0.96; R: 0.94; W: 0.93; D_m: 0.87; lgN_w: 0.78), but also lower absolute biases and reduced root mean square errors. (3) Simulations of Z and differential reflectivity (Z_DR) based on the REP DSD showed better agreement with XPOL observations, with biases of −1.14 dB (−3.04 dB for the AVG data) and 0.02 dB (0.12 dB for the AVG data), respectively. (4) The vertical resolution of the MRR had a significant impact on the retrieval performance. For stratiform rain, the MRR-retrieved DSD and rainfall parameters at resolutions of 30 m, 100 m, and 200 m showed high consistency with 2DVD measurements. For convective rain, however, the optimal agreement was observed at the 30-m resolution. Meanwhile, the performance at a resolution of 200 m was marginally better than that at 100 m. In summary, these results demonstrate that applying de-aliasing and vertical wind correction to MRR raw spectra significantly enhances MRR’s capability to retrieve the DSD and rainfall parameters for both stratiform and convective rain types, providing a more reliable technical means for investigating the fine-scale vertical microphysical structure of precipitation.