Abstract: To further improve the forecasting capabilities for severe convection, with multi-source observational data, ERA5 reanalysis data, and numerical model outputs, a diagnostic analysis of an extreme severe convective event influenced by a multi-cell severe storm in the central and northern regions of Guizhou Province from the afternoon to the night of May 5, 2023 was conducted. The results indicate that the multi-cell strong storm developed under atmospheric conditions characterized by dry upper layers, moist lower layers, strong thermal instability, and significant vertical wind shear. Radar echoes revealed multiple compact and intense isolated convective cells, one of which evolved into a supercell that produced extreme thunderstorm winds and large hail, exhibiting mesoscale features such as a weak-echo region in the lower levels, an echo overhang in the mid-to-upper levels, a mesocyclone, and high vertical liquid water content. The storm structure also displayed a pre-storm low-pressure area and a thunderstorm high-pressure area. In dual-polarization radar cross-sections, the descent of hail particles to the ground and a rapid decrease in the reflectivity factor were observed. Surface extreme winds were primarily caused by the gravitational drag of precipitation particles, raindrop evaporation, and hail melting. Regarding the convective triggering mechanism, the maximum center of the vertically integrated horizontal convergence (M-term) correlated well with the initial locations of convective cells. The triggering was mainly due to the coupling of surface mesoscale convergence lines and the 850 hPa wind shear. Discrepancies existed between subjective and objective forecasts for this severe convective event. Strengthening the verification and evaluation of numerical models, along with applying corrections and quantitative analyses based on observations, can effectively enhance forecasting accuracy.