Severe non-point source pollution has widely resulted from the nitrogen losses in paddy-field drainage in southern China, due mainly to excessive application of chemical fertilizers and unreasonable irrigation. The goal of this study was to improve the water use efficiency, while mitigating the reactive nitrogen losses in paddy fields. A controlled drainage system (CD) was designed to combine the subsurface pipe and open drainage ditches, with an open-ditch controlled drainage system (OD) as a control group. Specifically, the CD system consisted of a controlled drainage ditch and three field plots (CD1, CD2, and CD3) with a controlled subsurface pipe. By contrast, the three-field plots (OD1, OD2, and OD3) were free subsurface pipes in the OD system, where the field water freely drained into the drainage ditch through lateral infiltration. The drainage intensity and nitrogen concentration were monitored in various forms at the outlets of subsurface pipe and open ditch with a high frequency in six selected irrigation-induced drainage events, including three irrigation-drainage events with the fertilization (F1, F2, and F3) and three irrigation-drainage events without fertilization (D1, D2, and D3). The results showed that the drainage loss induced by six irrigation-drainage events accounted for 44.0% of the total amount of irrigation water in the OD system, indicating low water use efficiency. The combination of controlled drainage between the subsurface pipe and the open ditch greatly changed the drainage from the paddy field to the open ditch in the CD system. In all irrigation-drainage events except F3, the start time of open ditch drainage was later than that of subsurface pipe drainage, whereas, the peak of the intensity in the open ditch drainage was synchronized or significantly later than that of subsurface pipe drainage. Furthermore, the drainage peak and intensity of the open ditch in the CD system were much lower than those in the OD system among all six irrigation-drainage events. Specifically, the drainage peak of the latter was 1.3 to 8.2-fold that of the former, where the average drainage intensity was 1.5 to 4.4-fold. Compared with the OD, the CD decreased the drainage peak, intensity, duration time, as well as total drainage loss, where the proportion of drainage amount in the total amount of irrigation water dropped to 24.4%, indicating an effective role in drainage mitigation. In the irrigation-drainage event F1, F2, and F3, the concentrations of ammonium (NH4+) and total nitrogen (TN) in the drainage from the outlets of subsurface pipe (CD1, CD2, and CD3) and open ditches (CD and OD) gradually increased over time until the end of the drainage. Nevertheless, the concentrations of NH4+, nitrate (NO3-), and TN in the drainage from these same outlets gradually increased over time in the irrigation-drainage events D1, D2, and D3. Furthermore, the average concentrations of NH4+, NO3- and TN in the drainage from F1, F2, and F3 were much higher than those from D1, D2, and D3, indicating that the nitrogen loss was effectively reduced during the drainage management in a certain period after fertilization. As such, the open ditch in the CD system significantly intercepted a large number of nitrogen loads from subsurface pipe drainages. The nitrogen losses in the forms of NH4+, NO3- and TN from open ditch drainage in the CD system greatly decreased by 42.6%, 70.7%, and 39.3%, respectively, compared with the OD system. Consequently, the CD system can be expected to significantly reduce drainage loss and control non-point source pollution. This finding can also provide promising drainage control for the paddy field in southern China.