To gain information concerning the ecotoxicity of isoproturon (IPU) on aquatic ecosystems, six experimental ponds of 5 m3 each were studied. All the experiments were conducted during the summer over two years. Three different types of ecosystems were tested in 1994 and one type of ecosystem was selected and repeated in 1995 with three replicates. In each case, the initial concentration of IPU contamination was set at 10 microg/L. The IPU concentration was determined in the water column and in different species (mainly plants) of the microcosms. A first-order kinetic decrease in IPU concentration was observed in 1994, with half-life ranging from 15 to 35 d, depending on the microcosms. This relatively fast decrease was also confirmed in 1995, but it reached a constant value after two months. A high variability of the IPU concentration was observed in exposed plants, with bioconcentration factors ranging from 100 to 1,200 with large coefficients of variation. The observed plant bioconcentration factors are higher than those predicted by usual numerical models, probably due to the specific binding of IPU on one protein of the photosynthetic apparatus. Our data show that bioconcentration does not occur in mollusks but is important in photosynthetic organisms. Plant bioconcentration and microbial biodegradation are the main processes involved in the IPU decay in our outdoor aquatic microcosms.