In the Stimulation Program System-oriented Ecotoxicological Research (SSEO) three sites in The Netherlands were investigated for field effects of the grey veil of pollutants. At each site several studies were performed in order to arrive at an adequate weight of evidence and to improve causal inference of pollutant effects. This paper contains a synthesis of results of the studies, performed at one of the sites, the Demmerikse polder. This site is characterized by an anthropogenic layer of soil (in old Dutch: 'toemaakdek') on top of the natural peat. Lead, copper and zinc concentrations were elevated, with lead concentrations above a Netherlands environmental quality criterion (Intervention Value) in 66% of the samples. Issues discussed in the paper are: the sampling strategy, selection of maximum gradient and suitable community end-points, both in space and in time. Specific emphasis was given to causal inference of ecological effects of pollutants, related to direct versus indirect effects, functioning of ecosystems, normal operation range and risk assessment. The plausibility of metal effects could be demonstrated on a number of occasions. In the Demmerikse polder changes in the bacterial and nematode communities could be related significantly to metal concentrations and separated from other environmental variables, such as organic matter content and pH.