Mathematical models and computer simulation codes designed to aid in hazard assessment for environmental protection must be verified and validated before they can be used with confidence in a decision-making or priority-setting context. Operational validation, or full-scale testing via an 'appeal to Nature' in realworld situations, is usually the most ambiguous and least satisfactory part of the validation process. In most published studies, objective validity criteria are lacking and evaluation of the model is wholly subjective. This need not be the case, however, because acceptable model performance can usually be defined using relatively uncomplicated accuracy and precision criteria. Under these conditions, objective statistical tests can be formulated and executed to provide unambiguous proofs of validity in individual case studies. Such validation cannot, of course, demonstrate the global validity of a model; they merely provide a single instance of a failure to invalidate. The accumulation of a series of such validations, however, can give model users confidence in the general reliability and veracity of their decision tools.