Virtual models (vMs) are computer representations of biological processes (at various levels of organization) used to resolve complex system behaviors and emergent properties that cannot be understood solely from studying the individual parts. Computational power is now available to develop vMs capable of supporting predictive toxicology, and of reducing the dependence on in vivo animal studies for basic research and risk assessment purposes. The ultimate goal is to simulate in vivo adaptive responses to environmental change, and to predict the affects of various perturbations on system behaviors. Examples of vMs are presented from research in the fields of physiology, pharmacology, toxicology, and risk assessment.