Objectives: Chemosensory effects of stimulation by a chemical can either be irritating (trigeminal stimulation) or odorous (olfactory stimulation) or both. For odorous irritants, a clear-cut distinction between odour and irritation is difficult to make. The differences in the lowest concentration found to be irritating to the respiratory tract in humans when compared to experimental animals has brought much debate in the process of setting occupational exposure limits (OELs) for such chemicals. In this paper it will be discussed as to how far subjectively measured sensory irritation threshold levels can be used to establish OELs. Methods: Data on respiratory irritation of four odorous irritants were retrieved from public literature and discussed, viz. acetone, formaldehyde, furfural and sulphur dioxide. Results: Objective measures of irritation yielded results that differed from subjective evaluations. Important factors modulating the reported levels of irritation and health symptoms include the perception of odour intensity, exposure history and the individual's bias to report irritation on the basis of his/her prejudice or knowledge of the compound. Conclusions: Subjective measures alone are less appropriate for establishing sensory irritation thresholds of odorous irritants and are, therefore, less suitable to establish OELs without supporting evidence. Objectively measured irritation in humans, the Alarie assay (an experimental animal test assessing the concentration that results in a 50% reduction of the breathing frequency) and repeated exposure studies in animals may be of help to study objective irritation. If subjective measurements are used to study sensory irritation, the study design should at least include: Measurement of both incidence and severity determined at several concentrations, an appropriate (0 ppm) control condition, preferably a non-irritant odorant control exposure, validated questionnaires and correct concentration measurements. © Springer-Verlag 2005.