Simple evapotranspiration models with few data requirements and Penman reference evapotranspiration for grass (Eref) limiting the actual evapotranspiration are often used to estimate chemical fluxes in ecosystems. The aim of this paper is to show how much the interception loss might exceed Eref in a wind-exposed Danish Sitka spruce stand and to demonstrate the importance of the evaluation of model performance, here represented by throughfall measurements and chemical fluxes. Precipitation, throughfall, soil moisture and soil water chemistry were measured monthly for 1.5 years. Model input was daily precipitation, Eref, leaf area index, root distribution and plant available water. The model interception loss was calculated using an empirical relationship between precipitation and interception loss that was calibrated against the measured interception loss. Interception loss was found to be unusually high, on average 58% of precipitation, which was supported by measurements from two other years with interception losses from 62–68% of precipitation. Transpiration and evaporation from soil were constrained by Eref. The modelled percolation was compared to percolation calculated by a chloride mass balance based on measured values. The percolation was only about 20% of precipitation due to the exceptionally high interception loss, which exceeded Eref by almost 50%. Therefore, in wind-exposed forest stands the interception loss should be modelled separately, as done here, and be calibrated on measured values.