The degradation of soils due to various anthropogenic stress factors is alarming. Although chemicals are a major reason for soil degradation, most ecologists are not interested in studying such effects. We try to wake their interest by addressing a number of unsolved soil ecotoxicological problems that are related to disturbance ecology, biodiversity, ecosystem functioning and modelling. Features distinguishing chemical from natural stress render promising new aspects in disturbance ecology. Ecotoxicological studies are ideal models of disturbance, particularly regarding frequency, intensity or multitude of stress. Patterns of secondary succession after a major chemical damage can directly be related to the intermediate disturbance hypothesis. More knowledge on altered life history patterns following stress could support both evolutionary ecology and risk assessment. We raise the question if inherent resource competition makes communities more vulnerable to stress. Three aspects of ecotoxicological risk assessment are introduced: (1) exposure and bioavailability, which is directly connected to environmental heterogeneity; (2) tests on ecosystem functioning, suffering from major drawbacks; and (3) modelling. Here, promising approaches exist but need substantial input for being applicable to soils. Ecological modelling should put more emphasis on simulating both natural and chemical disturbances, including behavioural aspects and environmental variability. Finally, research needs for ecological risk assessment in soils are derived such as a simple system to assess the impact of chemicals on soil biodiversity, the inclusion of behavioural changes of keystone species or the consideration of density-dependent effects. Common research efforts of basic ecologists and soil ecotoxicologists could render a lot of mutual benefits.