The study evaluated stress-related complaints of residents living on contaminated ground. Two hundred and fifteen residents of an area that was contaminated by toxics of a former coking plant (toxins were benzo(a)pyrene, dicyclopentadiene, benzol, toluol, and xylol) were compared to a control group composed of 200 subjects living in an uncontaminated part of the same city. Data are based on a standardized interview concerning stress-related somatic and psychological symptoms and on exposure parameters. The residents of the contaminated area were also required to report their estimation of danger and trust in the city council's managing of the redevelopment. The results show that the residents reported significantly more stress symptoms than the control group. In a stepwise regression analysis, symptoms turned out to be best predicted by cognitive variables. Exposure parameters were excluded from the regression model because they provided a comparatively small contribution to the explanation of variance.