A rat bioassay has been developed to provide an objective approach for the identification and classification of upper and lower respiratory tract irritants, with particular emphasis on the concentration-dependent induction and regression of lesions characteristic of asthma, such as persisting non-specific airway hyperreactivity, inflammation and ensuing mismatch of the ventilation-perfusion relationship. For the identification of respiratory allergy, the established guinea-pig bioassay has been further refined. Refinement focused on procedures making this animal model more robust to changes in study design. Attempts were made to allow differentiation of non-specific and specific bronchial hyperresponsiveness and to minimise the use of hapten-protein conjugates for elicitation of respiratory allergy. It appears that the combined assessment of specific pathologic features such as airway eosinophilia and the evaluation of several breathing parameters during hapten and acetylcholine bronchoprovocation challenge make it easier to distinguish effects caused by irritation and respiratory hypersensitivity. Findings support the conclusion that current guinea-pig models require specific optimisation of sensitisation and challenge procedures for each chemical class tested.