Mathematical models (should) play a central role in quantitative research, both in the design of experiments and in the analysis of their results. This also holds for research on stress on individual organisms, where stress is defined as an environmentally induced change in their (eco-physiological) behaviour, implying the necessity to know the behaviour in absence of stress in some detail. The individual can effectively be modelled in terms of a dynamical system, where stress shows up as a change in one or more parameters that control the behaviour of the system. After a more detailed presentation of the empirical cycle and an introduction to dynamic systems, I will discuss this approach in the context of generalised ecotoxicity, where presence (e.g. toxicants) or absence (e.g. dioxygen) of particular chemical compounds in the environment might affect a variety of endpoints (feeding, growth, reproduction, maintenance, survival). To this end I will discuss chemical transformation in the environment (speciation, ionisation, degradation, absorption), transport to and from the individual (various uptake and elimination routes, popular transport models), metabolic transformation, effects of nutritional status on kinetics and effects (lethal and sublethal).