To assess the validity of peak expiratory flow (PEF), measured by means of a peak flow meter, in comparison to forced expiratory volume in one second (FEV1) for measuring the airway response during specific bronchial challenges, we registered PEF and FEV1 in a random sequence during 75 positive (decrease in FEV1 of > or = 15% from baseline) and 75 negative (decrease in FEV1 < 15% from baseline) challenges with chemicals or allergens. The correlation between PEF and FEV1 in terms of absolute values and of percentage of change from baseline was statistically highly significant (p < 0.001) in all challenges and in the different pattern of response, immediate, dual and late. Relative operating characteristic analysis showed that an absolute decrease in PEF of > or = 70 l.min-1 in the immediate and of > or = 80 l.min-1 in the late phase of the response (cut-off points) gave optimal discrimination between challenges with a bronchoconstrictive response (defined as a FEV1 decrease > or = 15% from baseline), and challenges without. These cut-off points, however, were highly specific (92 and 93.3%, respectively), but not as sensitive (70.6 and 61.1%, respectively), and smaller absolute changes in PEF from baseline do not exclude a bronchoconstrictive response. Our data suggest that PEF readings are a useful diagnostic tool in assessing the airway response during a specific bronchial challenge in asthmatics, although with some limitations. PEF readings can be used to monitor the late response to a challenge during the evening and the night if PEF readings are carefully considered in relation to clinical symptoms.