Drip irrigation of potatoes can be an effective way to improve the water and nitrogen use efficiency in the soil and hence to reduce the environmental pollution. The experiment was carried out in lysimeters on two different soil types: coarse sand and sandy loam. For both soil types, plots were both irrigated using a drip line system and an overhead system. An automatic roof was used to exclude the lysimeters from natural precipitation. Climatic data from a nearby climatic station was used for calculating the FAO56 reference evaporation, which was subdivided into potential evaporation and potential transpiration to drive the HYDRUS-2D model. This subdivision was based on three different additional models: The irrigation system, MARKVAND, the soil plant atmosphere system model, Daisy, and a newly developed decision support system, DSS. Outputs from the HYDRUS-2D model were compared with the volumetric water content in the soil measured by TDR probes installed parallel to the tillage direction at different positions in the potato ridge. Daisy and the DSS seemed to simulate the crop development best and it was therefore expected that the two models would estimate the potential evapotranspiration most correctly. However, outputs from the HYDRUS-2D model were most favorable when using the third model, MARKVAND, for the subdivision of the potential evapotranspiration. It seems therefore that the relation between crop development parameterization and the crop coefficients needs to be further investigated. In general, HYDRUS-2D estimated soil water content well for both soils types when modeling the drip irrigated plots. On the contrary, results showed large deviations from the measured water contents when simulating the overhead-irrigated plots especially for the sandy soil. The large deviations were probably related to the hydrophobic nature of the soil leading to a runoff of water from the side of the ridges to the “ridge valleys”.