Climate change-induced extreme precipitation causes coastal flooding. A streamflow simulation in coastal watersheds, Wolf River Watershed (WRW) and Jourdan River Watershed (JRW), was conducted using the Soil and Water Assessment Tool (SWAT) to compare variation in flow at different climates and to analyze the flood frequency. Baseline models were auto-calibrated with SWAT calibration and uncertainty programs (SWAT-CUP). Kling–Gupta efficiency (KGE), defined as the objective function in SWAT-CUP, ranged from 0.8 to 0.7 in WRW and from 0.55 to 0.68 in JRW during the calibration–validation process. Results indicated reliability of the model performances. Monthly averaged baseline flow was 1% greater than historical and 8.9% lower than future climate in WRW. In JRW, monthly averaged baseline flow was 11% greater than historical and 5.7% lower than future climate. Flood frequency analysis showed the highest 1% exceedance probability in annual maximum series (AMS) of baseline model in WRW, whereas AMS of projected model was estimated the highest in JRW. This study aids in preparing for future flood management.