In this paper a new method is introduced to assess the impact of storminess during the pre-instrumental period. The method is based on written sources. For this purpose Flemish documentary sources are taken as an example originating from a polder region between Ghent and Antwerp dating from the period 1488 to 1609. These documentary sources are continuous and homogeneous. The polders were protected by dikes, which had had to be maintained regularly. The annual upkeep was answered for in the annual dikes accounts in which nearly each storm that caused damage to the dikes was recorded. In order to assess the impact of storminess, these storms have been categorised, based on the following three principles: (a) the exact wording in the text, (b) the circumstances in which a storm appeared, (c) the consequences to which a storm has led, e.g. number of acres flooded. Accordingly eight separate categories are distinguished with storms ranging from one point to eight points. So in the first category the least important storms are given only one point. In category eight the storm surges appear, that have led to the flooding of large areas. These storms are given eight points. As the method enables us to assess the impact of storminess throughout the period, it shows only a small increase of storminess during the first quarter of the sixteenth century and a large increase in storminess during the second half of that century. Finally, looking at the storm pattern at Flushing between 1848 and 1990, there also appears to be a relative strong increase in storminess beginning around the 1950s. Whereas such a similarity in increase in storminess between these two quite different periods can hardly be coincidental, there must be some common cause. Could it be that periods of a relative strong increase of storminess precede periods of climatic change?