The development of an integrated theory of subsurface drainage based on hydrology and hydrogeology concepts is presented. The historical context, the main hypothesis derived from the Boussinesq equation and the validation of the model predictions are discussed. Theoretical developments of this equation demonstrate that a single parameter (σ)-a combination of soil and drainage system properties-is sufficient for predicting the dynamics of subsurface drain flow rates. We also demonstrate that these drain flow rates are a function of the level of water replenishment in the system (classically the water table elevation), of the recharge intensity of the aquifer and of a buffer function related to the swelling or deflation of the water table shape during recharge events. For values of σ > 1, the buffer role of the water table is negligible. In that case approx. 13% of the water table recharge contributes to the flow rate, which is shown to explain the observed disconnection between water table elevations and peak flow rates at the outlet of classic agricultural drainage systems and to predict these peak flow rates accurately. A modelling approach based on this theory and validated experimentally (SIDRA model) allowed us to test the quality of the peak flow prediction. The SIDRA model also includes a surface runoff module and has been coupled to different modelling tools and used to analyse the impacts of subsurface drainage on water quality. The approach contributed towards the development of tools that helped to connect better the drainage systems to the hydrological functioning of watersheds.