Artificial drainage is a common agricultural practice to improve moisture and aeration conditions, but it may contribute to the diffuse pollution of surface water bodies. To evaluate the environmental impacts of artificial drainage, monitoring is necessary, and these monitoring programs need to be designed carefully to capture the underlying processes correctly. Using data from a monitoring program in northeastern Germany, we studied different methodological aspects concerning the design and evaluation of such studies. To unravel scale effects, a hierarchical monitoring approach is advisable. Even within apparently homogeneous fields, significant differences in the discharge and the NO3–N concentration patterns were measured. Sampling should not be restricted to one or two seasons, as there is a considerable interannual variability of the NO3–N losses. Pronounced diurnal dynamics may occur as well, and thus composite samples using automatic samplers instead of grab samples should be taken. Due to the high temporal variability and the short response time of artificially drained catchments, weekly to monthly sampling introduces large errors into the estimation of the total loads irrespective of the applied load calculation method. Overall, linear interpolation performed best but still tended to underestimate the true loads. Although sampling every third or fourth day might allow an adequate calculation of the total loads, a correct interpretation of the data might still be difficult. These uncertainties are very important not only when evaluating the environmental impact of artificial drainage but also when applying water quality models where input data uncertainty frequently remains unconsidered.