A large part of the Netherlands consists of polder area. Therefore the inhabitants have to rely on the Dutch flood defence system to protect them against floods. A vast network of drainage canals ensures that the seepage water and the precipitation water is transported from the polders toward the outside water. There are several mechanisms which can cause failure of the embankments; one of the mechanisms is macro instability of the inner slope. The stability is mainly determined by the pore pressures inside the dike. They cause a reduction in effective stress which is necessary to stabilize the dike. Precipitation acts as a recharge of the groundwater and causes the pore pressures to rise; the exact influence of precipitation on the pore pressures is however not described in literature. The objective of the study was: to compose a method to assess the probability distribution of the phreatic surface that includes the effect of precipitation. With a probabilistic computation it is possible to assess how safe the region behind the dike is instead of only knowing if the area is safe enough. The method that is developed is based on finite element computations of multiple precipitation events for dikes consisting of different hydraulic soil characteristics. In the study the probability density function of the phreatic surface is related to statistical information of precipitation events in the Netherlands. The result from the study is the probability density function of the rise of the phreatic surface in 11 combinations of soil type and permeability. They show that precipitation can have a significant influence on the pore pressures. Each probability density function contains information on the rise of the phreatic surface with a probability of exceedance between 0,5 and 0,001. The rise of the phreatic surface ranges between several centimetres in a dike consisting of permeable sand, to full saturation in a dike consisting of medium permeable clay or peat. The most important finding with respect to the current assessment method (Technisch rapport waterspanningen bij dijken) is that precipitation causes non-hydrostatic pore pressures. The current guidelines prescribe hydrostatic pressures, that results in an overestimation of the pore pressures. This means the probability of failure will be lower than expected regarding to governing schematization method. The most important finding respective to the probabilistic method in ‘flood risk of regional flood defences’ is regarding the probability mass of the phreatic surface. In this thesis report is concluded that the failure due to a high or a low phreatic surface must be regarded separately. The probability mass of the spectrum of high phreatic surfaces is proposed to be 100% instead of 10% as is assumed in ‘flood risk of regional flood defences’. This means the computed failure probability will be larger.