Benthic animals are a key element of many marine and estuarine monitoring programs. They are, in contrast to e.g. plankton or fishes, sedentary and must adapt to environmental and human-induced disturbances or perish. They are immediately dependent on the quality of the overlying water and very sensitive to habitat disturbance. The benthos integrates the effects over a considerable period of time. Many fishes, birds and mammals depend directly or indirectly on the benthic fauna. However, monitoring the benthos is very costly involving ship-time and laboratory analysis that is very intensive. The time-consuming nature of the laboratory analysis leads to a lag phase between sampling and production of results. Therefore, the most appropriate methods should be used for both sampling the benthic fauna and analysing the data in the light ofthe objectives of the monitoring program. During the last 10 years, I have been involved in several inventorying and monitoring studies in the Netherlands. In this dissertation part of the macrobenthos data recorded during these surveys are re-analysed in the light of some problems encountered when evaluating biological changes in the marine environment. Chapter 2 gives a detailed description of the surveys and the macrobenthos data used in chapter 3 to 8. All surveys have been carried out in the last fifteen years in the North Sea and adjacent estuaries in the Southwest of the Netherlands. To allow proper data management ofthe huge amount of data resulting from these surveys, the data have been stored in a relational database. In chapter 2 the structure of the database and some problems encountered during the design and maintenance are reported. Chapter 2 finally gives an overview of the multivariate techniques used in the other chapters. Multivariate methods of data analysis describe the variability of the structure of the assemblages as a whole. In contrast to univariate methods, such as the analysis of variance, they take into account the covariance among the descriptors ...