AbstractSpatio‐temporal variations of nitrate‐nitrogen ( ‐N) leaching is driven by both soil hydrology and biogeochemistry. However, the widely used soil hydrology and biogeochemistry models have their weaknesses in simulating soil N cycling and soil water movement processes, respectively. In this study, we proposed an alternative approach by simply combining the HYDRUS‐3D and DNDC models to investigate the spatio‐temporal variations of ‐N leaching on a representative tea garden hillslope in Taihu Lake Basin, China. Results showed that the soil hydrology and N cycle were well simulated by HYDRUS‐3D and DNDC models, respectively. Based on the leaching equation, the soil water flux simulated by HYDRUS‐3D and soil ‐N content simulated by DNDC were combined to calculate the leachate ‐N concentrations with good accuracy. The accumulative ‐N leaching flux during the simulation year was 71.7 kg N ha−1, with remarkable spatio‐temporal variations on this hillslope. Hot spots of ‐N leaching were observed in blocks 24, 27, 31, 34, 37, and 40 with accumulative leaching fluxes > 82.0 kg N ha−1 y−1. The spatial variation of ‐N leaching was mainly controlled by soil texture and soil hydraulic properties. Hot moments of ‐N leaching were observed after the applications of spring fertilizer (16 March) and basal fertilizer (30 October). The temporal variation of ‐N leaching was mainly controlled by precipitation and the spring fertilization. Methods and findings of this study will be benefit for the risk assessment of non‐point source N loss and the precise agricultural management.