Rapid damage prediction for wind disasters is significant in emergency response and disaster mitigation, although it faces many challenges. In this study, a 1-km grid of wind speeds was simulated by the Holland model using the 6-h interval records of maximum wind speed (MWS) for tropical cyclones (TC) from 1949 to 2020 in South China. The MWS during a TC transit was used to build damage rate curves for affected population and direct economic losses. The results show that the Holland model can efficiently simulate the grid-level MWS, which is comparable to the ground observations with R2 of 0.71 to 0.93 and mean absolute errors (MAEs) of 3.3 to 7.5 m/s. The estimated damage rates were in good agreement with the reported values with R2 = 0.69–0.87 for affected population and R2 = 0.65–0.84 for GDP loss. The coastal areas and the Guangdong-Hong Kong-Macao Greater Bay Area have the greatest risk of wind disasters, mainly due to the region’s high density of population and developed economy. Our proposed method is suitable for rapid damage prediction and supporting emergency response and risk assessment at the community level for TCs in the coastal areas of China.