We have measured bronchial responsiveness (BR) to histamine on two occasions between 5 and 24 h apart, to determine if conventional and new indices of BR are repeatable. A random sample of 29 healthy male subjects with a mean age of 19 (SD 3.44) years from a larger study repeated a Yan method test of BR, recording both partial and maximal expiratory flow volume (PEFV and MEFV) curves. From the MEFV curves log-dose slopes (LDS) for forced expiratory volume in 1 sec (FEV1), forced expiratory flow between 25% and 75% of forced vital capacity (FVC) (FEF(25-75%)), mean expiratory flow at 30% and 40% of FVC (MEF30, MEF40), and the first moment of the spirogram (alpha1) truncated at 75% and 90% of FVC were calculated, as well as the provocative dose that induces a 20% fall in FEV1 (PD20FEV1). From the PEFV curves LDS for alpha(1)75% and alpha(1)90%, and MEF30 and MEF40 were derived. Apart from MEF30 and alpha(1)90% the second test was significantly lower (P<0.05) than the first when measuring the repeatability of spirometric indices, whereas the LDS of the indices showed no significant change. The repeatability expressed as intra-class correlation coefficient (ICC) was highest for LDS FEV1 (0.87), second highest for LDS MEF40 (0.67) and LDS MEF30 (0.65). The LDS for moment indices were much less repeatable and the lowest ICC was found in all LDS indices derived from PEFV curves. Within-subject variance was not influenced by atopic status, smoking habits or recordable PD20FEV1. As tests for bronchial hyper-responsiveness (BHR) the LDS of FEV1, MEF40 and MEF30 seem to be acceptable for use in population studies.