Abstract:[Objective] Runoff responses to climate changes in Lanjiang River basin were analyzed in an effort to provide a scientific basis for improving sustainable water resources, flood control, and drought resistance.[Methods] The applicability of the HEC-HMS hydrological model in this basin was determined by calibrating the processes of daily rainfall and runoff, in addition to six processes of rainstorms and floods from 2015 to 2018. Based on the statistical downscaling model (SDSM), the climate data of three climate change scenarios (RCP 2.6, RCP 4.5, and RCP 8.5) under the CANESM2 model from 2030 to 2100 were downscaled for generating future daily precipitation time series at six meteorological stations in Lanjiang River basin in order to predict the runoff in response to future climate changes.[Results] The average correlation coefficients of the HEC-HMS model for floods and daily runoff simulations were 0.89 and 0.77, respectively. The average efficiency coefficients of the HEC-HMS model for floods and daily runoff simulations were 0.86 and 0.76, respectively. Precipitation in the study area under the RCP2.6 scenario decreased by 0.82% compared with the base period (2015-2018), and increased by 6.18% and 18.17% under the RCP4.5 and RCP8.5 scenarios, respectively. Average annual runoff under the RCP2.6, RCP4.5, and RCP8.5 scenarios increased by 17.00%, 26.22%, and 41.93%, respectively, compared with the base period.[Conclusion] The HEC-HMS model is applicable for runoff simulation in the Lanjiang River basin. In the future, runoff in the Lanjiang River basin will show a significant upward trend, and increase synchronously with the enhancement of radiation forcing. When radiative forcing increases to 8.5 W/m2, runoff will increase by 49.49 m3/s every 10 years, and predictably, the average annual runoff will reach 1 101 m3/s at the end of the 21st century. Additionally, annual runoff will fluctuate sharply, runoff in the flood season will account for a higher proportion of the year, and drought and flood events will tend to be frequent.