There is widespread belief that methacholine responsiveness in asthmatic subjects is closely related to histamine responsiveness, and that the two agents may be used interchangeably in the measurement of non-specific bronchial responsiveness (NSBR). Because this view has been challenged, we have examined the repeatability of measurements of bronchial responsiveness to methacholine and histamine and the relationship between them, in groups of 20 adult asthmatic subjects. Bronchial responsiveness was expressed as the cumulative dose (in both micrograms and mmols) provoking a 20% fall in FEV1 (PD20). The predicted 95% confidence limits for the second PD20 measurement of a further pair were within the range 0.5-2.0 x first PD20 for both agents. When the two agonists were compared in the same subjects, a significant difference in potency was noted (ratio of geometric means PD20.methacholine to PD20.histamine 2.19 [micrograms], 3.43 [mmols]; p = 0.0003). Furthermore, the variance of the differences of the pairs of log PD20.methacholine and PD20.histamine measurements was found to be significantly greater than that of either the paired methacholine measurements or the paired histamine measurements (p less than 0.01). We conclude firstly that methacholine is a less potent bronchoconstrictor than histamine and secondly that, while inhalation tests with either agent broadly reflect the degree of NSBR, they measure different phenomena and cannot be used interchangeably even after allowance is made for the difference in potency.