Model simulations rely on rainfall measurements, mostly recorded by point measurement devices, i.e. raingauges. When using data from extreme rainfall events, the model user has to take into account the validity of point measurements in space. The presented study outlines investigations in the field of rainfall variability due to the actual situation of general weather types during several events. The Boye catchment is part of the Emscher system (850 km2), located in north west Germany. Data from seven continuous raingauges (1990 to 1996) were analysed, together with data from synoptic classifications of the general weather type. Various statistical parameters were computed for the 200 most important rain events, both stationwise and for the whole event over the complete catchment area. The investigation results show that for particular synoptic rainfall situations, the raingauge in the centre of the catchment overestimated the areal rainfall calculated by the Thiessen procedure from all seven raingauges. The results also demonstrate the spatial variation of rainfall as well as observed event duration and event volume variation for different synoptic situations. It is pointed out that reduction factors should be applied to the data of the single raingauge in the centre of the 80 km2 catchment area used in longterm simulation.