Bronchial responses to occupational sensitizers measured by peak expiratory flow rate (PEFR) and forced expiratory volume in 1 s (FEV1) during late reactions (between 90 minutes and 8 hours after exposure) were compared in two groups of 88 subjects who had undergone specific inhalation challenges in the laboratory. The first group had what was considered a positive reaction (a fall of at least 15 percent in FEV1) whereas the second group's reaction was interpreted as negative (fall in FEV1 less than 15 percent). Although the correlation in terms of percentage of change from baseline values was statistically significant, the correspondence was poor. PEFR proved far less sensitive than FEV1 in detecting a reaction. Whereas the mean maximum change in FEV1 overall was 27 percent, the mean maximum change in PEFR at the same time interval was only 16 percent. Moreover, individual correlations between the percentage of change in FEV1 and PEFR were satisfactory (r2 greater than 0.80) in only 32/88 subjects (36 percent). No subject who was considered to have a negative challenge according to FEV1 had a change in PEFR greater than 20 percent. We therefore conclude that changes in PEFR are far less sensitive than changes in FEV1 in detecting responses during late reactions to occupational sensitizers.