AbstractThe environmental effects of stream regulation are often assessed with reference to changes in river benthos, and biotic scores and indices have been used to summarize data on faunal community structure and richness. Although they have their use in presenting complicated data in a simplified form, they often do not detect the more subtle effects of regulation. In lowland regions the regulation of rivers is most often due to water abstraction, channelization and the construction of weirs or sluices. The effects of this form of stream regulation may sometimes be particularly difficult to demonstrate with conventional biological assessment techniques. In this study we examine a length of stream regulated with sluice gates using conventional techniques and test the applicability of the mesohabitat method (categorization of reaches on the basis of their substratum types) in describing changes brought about by regulation. Conventional biological assessment techniques were unable to demonstrate the regulation effects. This was because, although changes had occurred along the reach, these were not associated with a significant change in scores and indices based on family richness. The changes were subtle and were most clearly revealed by clustering techniques and analysis of variance. Most of the change was due to altered physical habitat and this was reflected in the proportions of mesohabitats above and below the sluice gates. The association of mesohabitats with particular faunal assemblages allows an assessment of the overall effects of habitat change on the composition of the faunal community.