Although total heavy metal contents of soil are often used to express the degree of contamination, they are of little value to judge environmental effects. The main objective of this thesis was to develop and test methodologies with which environmental risks of cadmium pollution of field soils can be evaluated. A method to determine the mobile and bioavailable fraction of the total Cd content may be soil extraction with 0.01 M CaCl 2 . Chemical processes that control the cadmium activity in such an extract were quantified. This is relevant information when Cd contents in CaCl 2 -extracts need interpretation in terms of bioavailability and leachability.Cadmium behaviour at the field scale was studied using soil samples taken from an arable field in the Dutch Kempen region. Spatial variability of total Cd contents, CaCl 2 -extractable Cd contents, soil-pH, and organic matter content was distinct and different for all parameters. Spatial variability of soil-pH and organic matter content explained a major proportion of the variability of Cd contents.Due to spatial variability, a high sampling density was needed for reliable estimation of pollution boundaries. When polluted soil needs remedial action, reduction of the research effort by minimizing the sampling density may well lead to an overall increase in sanitation costs because the area where concentrations exceed a critical threshold level becomes larger.The long-term effect of soil heterogeneity on cadmium behaviour in soil was evaluated using a stochastic simulation model. Predicted uptake of Cd by barley and leaching of Cd to groundwater was much higher in a heterogeneous field as compared to an equivalent homogeneous field. Variability of soil hydraulic properties as well as soil chemical parameters caused these differences. This demonstrated the importance of soil heterogeneity for environmental impact assessment of soil contamination.