This study aims at describing the spatial distribution – and its temporal variations - of discarding intensity (i.e. expected weight of discards for a standard trawl haul) for the 6 main species discarded by the Dutch beam trawl fisheries. For each species, the spatial distribution (quarterly maps for the period 2013 to 2017) is estimated using statistical models that take spatial and temporal correlation into account, which also allowed to test for the effect of a number of factors related to geography, environment, fishing practices and operational aspects on discarding. The data used to fit those models came from the observer trips and self-sampling program conducted at Wageningen Marine Research and from discards sampling trips conducted by the fishing industry. As by-product, the models provide descriptors of the temporal and spatial scales at which the discards of a given species are structured. The distribution of the expected discards per haul for dab was highly variable from quarter to quarter, with generally high discarding intensity in front of the southern coast of the Netherlands in quarter 1, a discarding intensity which is high on the German bight and low in front of the Dutch coast in quarter 3, and variable distributions for quarter 2 and 4. For plaice, the distribution was more stable, with high values consistently observed in the south of the area (between the south of the Netherlands and England), with occasional hot spots on the German bight. For sole, discards were not observed on the north-western part of the area, and a hotspot of sole discarding was found consistently in front of the southern coast of the Netherlands, occasionally expanding towards England or to the northern coast of the Netherlands. Discarding of turbot first occurred with a low intensity along the coast from Belgium to Germany. After the fourth quarter of 2015, high discarding started to occur, first limited to the small area in the southern North Sea, but progressively expanding to a larger area in the southern and ...