In the context of climate change, the stakes surrounding water availability are rapidly intensifying. Decomposing and quantifying the effects of climate on discharge allows us to understand their impact on water resources better. We propose a methodology to separate the effect of change in the annual mean of climate variables from the effect of the intra‐annual distribution of precipitation. It combines the Budyko framework with Land Surface Model (LSM) outputs. The LSM is used to reproduce the behavior of 2134 reconstructed watersheds across Europe between 1902 and 2010, with climate inputs as the only source of change. We fit a one‐parameter approximation of the Budyko framework to the LSM outputs. It accounts for the evolution of the annual mean in precipitation (P) and potential evapotranspiration (PET). We introduce a varying parameter in the equation, representing the effect of long‐term variations in the intra‐annual distribution of P and PET. To better assess the effects of changes in annual means or intra‐annual distribution of P, we construct synthetic forcings fixing one or the other. European results show that the trends in the annual averages of P dominate the trends in discharge due to climate. The second main climate driver is PET, except over the Mediterranean area, where changes in intra‐annual variations of P have a higher impact on discharge than trends in PET. Therefore the effects of changes in the intra‐annual distribution of climate variables are to be addressed when looking at changes in annual discharge.