Some hydrologic models oversimplify crop growth simulation. Although some studies coupled hydrologic models with crop models, there are several hypotheses and unconsidered factors in these studies, such as no irrigation. To improve simulation of discharge, soil moisture, and evapotranspiration, the Variable Infiltration Capacity (VIC) model was coupled with a crop growth model, incorporating an improved evapotranspiration module based on the soil water stress method. The new evapotranspiration module improved two water stress errors of the Environmental Policy Integrated Climate (EPIC) model. Considering the impact of agricultural irrigation and crop rotation, regional crop growth and hydrology were simulated in the upstream watershed of the Qingkou River in the east of China. Results of simulation by the Coupled Hydrology-Crop growth (CHC) model were compared with observed discharge data, measured soil moisture data, and evapotranspiration data obtained from remote sensing inversion. The daily Nash efficiency coefficient was 0.78 for the calibration period from 1978 to 2008 and was 0.76 for the verification period from 2009 to 2016. The simulated soil moisture was found to be significantly correlated with observed soil moisture, better than the simulation by the VIC model alone. The daily efficiency coefficient of the CHC model simulation increased by 0.13 for the verification period compared with that of the VIC model; and the transpiration module of the CHC model had the best simulation accuracy among the VIC, EPIC, and CHC transpiration modules. Combined with observed data, the reason for the water stress error of the EPIC model, reported by some references, was found: A parameter needs to be further adjusted combined with the actual condition instead of being constant. Thus, the coupled hydrology and crop growth model improved the simulation of hydrological elements and can help guide the efficient use of water resources and adjustment of agricultural cropping practices.