This paper analyses the spatial distribution of fishing effort in a sample of 25 Dutch commercial beam trawlers fishing for sole and plaice in the period 1993-1996, based on an automated recording system with an accuracy of about 0.1 nautical mile. Intensive fishing occurred along the borders of the closed areas (12 mile zone and the 'plaice-box', a protected area in the eastern part of the North Sea) and at certain off-shore grounds in the southern and central North Sea. Effort distribution was studied within 30 x 30 (ICES rectangles), 10 x 10, 3 x 3 and 1 x 1 nautical mile squares and showed a patchy distribution. The degree of patchiness decreased with resolution. Within 3 x 3 mile squares, beam trawling was randomly distributed in some parts of the most heavily fished ICES rectangles but patchily distributed in others. Within 1 x 1 mile squares, the distribution became random within more than 90% of the squares. The micro-distribution showed a remarkable similarity between the 4 years with a mean coefficient of overlap of 0.66, range 0.56-0.76. The microdistribution of the sampled vessels was raised to the total Dutch fleet in order to estimate the frequency at which the sea bed was trawled. It was estimated that during the four year study period in eight of the most heavily fished rectangles of the North Sea, 5% of the surface area was trawled less than once in 5 years and 29% less than once in a year. The surface area of the sea bed that was trawled between 1 and 2 times in a year was estimated at 30%. The surface area trawled more than five times in a year was estimated at 9%. The relevance of the findings for the study of the impact of beam trawling on the benthic fauna is discussed.