Microseismic events or acoustic emissions induced by petroleum extraction in the Ekofisk field operated by the Phillips Petroleum Licence 018 group in the Norwegian sector of the North Sea' were monitored for 18 days in April 1997. The monitoring consisted of a 6 level, triaxial VSP wireline tool deployed within the reservoir in the 2/4-C 11 a observation well located near the crest of the field. 2100 microseismic events were recorded, corresponding to roughly 5 per hour. The rnajority of the events were located in the upper part of the reservoir, predominantly in low porosity layers overlying relatively porous layers which are undergoing water flooding and compaction. Events were most accurately located near the monitoring borehole, with the position of events further away being more uncertain. A cluster of activity surrounded the nearby well C 16, confirming the event positioning accuracy. Within 300 m of the well, most events were precisely located to an accuracy of better than 30 m, and found to cluster along discrete lineations. These lineations, which are parallel to major structures in the region, are attributed to induced definition of preexisting fractures. Comparisons with simulations of the water flood front indicated that the saturated zones around the injection wells were aseismic. The position of the water flood front was not accurately enough known to test if any of the clusters of microseismicity were associated with it. However, compaction of the chalk may be strongest at the water front. If a significant proportion of the total acoustic activity is caused in association with the water front, then time-lapse microseismic monitoring could represent a useful tool for tracking the position of the water flood.