A drift experiment was carried out in February 1991, using waterfowl that had accidentally died in fishing nets on the IHsselmeer in previous years. A total of 383 birds was marked individually, using a combination of a numbered tag and some coloured plastic thread (193 Great Crested Grebes Podiceps cristatus, 101 Goldeneyes Bucephala clangula, 84 Red-breasted Mergansers Mergus serrator, 2 Tufted Ducks Aythya fuligula, 2 Goosanders M. merganser and 1 Scaup A. marila). The birds were thrown overboard on 6 Feb 1991 at eight different locations, between 2 and 70 km from the coast (figure 1). Unfortunately, the weather changed dramatically just a few days before the experiment. While the wind had been mainly westerly during the previous period, it changed to the east and increased to force 7B (figure 2), while temperatures fell markedly to well below zero degrees centigrade. Moreover, a lot of snow made the searching for dead birds on the beach very difficult from 10-18 Feb. Only seven birds were recovered, within 12 and 33 days (table 1). Six ducks that were put overboard at 6 km of the coast (G 6) landed on the nearest shore despite the strong easterlies. One that was dropped at 10 km offshore (S 10) landed on the island of Texel, oiled, after 33 days (170 km NNE). Birds that had stayed longer in sea than 28 days were in various stages of decomposition, those that were recovered after two weeks were still ’fresh and complete’ corpses. From most previous experiments it was concluded that wind was the main factor responsible for both drifting speed and direction. The computer model, which considered wind as the only drift factor, was not appropriate to predict the landing places correctly (figure 3). The same applied for models in which wind was only one factor, also considering currents and tide. When wind was totally ruled out, strandings were predicted from dropping place G 6 (within 24 hours) on the beach of Goeree. The most interesting result of this experiment was the conclusion that even during strong offshore winds, corpses may reach the coast. It is suggested that (some of) the corpses did not so much float after being dropped, but rather sank to somewhere in the water column. Hence, it is recommended to repeat the drift experiment, bul not before a ’floating’-experiment has been performed.