In agricultural areas, nitrogen (N) pollution of surface water depends mainly on agricultural practices and the hydrology of the catchment. The presence of subsurface drainage networks modifies the hydrological behavior and accelerates the transfer of this pollution. Modeling nitrate (NO3-) transfer requires the use of large and complex data sets, which are generally difficult to collect. The main objective of this study is to evaluate a modelling assumption that would only need the prewinter nitrate pool (PWNP) as an input for modelling nitrate losses. PWNP represents the part of nitrate not assimilated by plants or resulting from the mineralization part of organic matters during the preceding summer and autumn seasons. Generally, the PWNP is evaluated by soil samples and nitrogen quantification, which is expensive and time consuming. The PWNP simulation approach for nitrate losses was validated using two different models. First, the simulation was performed over the subsurface-drained catchment of Orgeval (East of Paris, France) using the semi-distributed HYPE model. Second, the simulation was performed at one subsurface drainage network of the Chantemerle (35 ha; Sub-catchment of the Orgeval) using SIDRA-N model. The HYPE model flow simulation was performed across observation data at different sub-catchments scales ranging from 2.5 to 7.1 km2. In this model, we consider that the nitrate losses are influenced by the transport dynamic in the soil and the denitrification capacity of soils and the stream. Results show a good agreement between model outputs and observations (Nash–Sutcliffe coefficient NSE of 0.75 and 0.7 for water flow and nitrate flux, respectively). Likewise, the comparison between calibrated PWNP and measured one from field survey (annual campaign managed by farmers) showed a significant positive correlation. In SIRA-N model, we use an equivalent PWNP, because this model does not consider the biological N-cycle in the soil. The application of this model at Chantemerle sub-catchment provides satisfactory results in terms of nitrate concentration and flux. The PWNP modeling approach used in hype and SIDRA-N models allows accurate simulations of nitrate losses at the scale of small subsurface drained catchment. This simplified modeling approach demonstrates that PWNP can be used as driving pollution factor for the evaluation of nitrate losses from drained agricultural catchments, and allows us to propose it for wider use in agriculture and its environmental impact.