Guidelines for asthma management advocate home peak expiratory flow (PEF) monitoring. It is commonly stated that PEF variability is a good proxy of bronchial hyperresponsiveness (BHR), a hallmark of asthma. However, this has hardly been tested longitudinally, as required to monitor asthma. This study assesses which PEF index correlates best with BHR longitudinally and whether the correlation improves when correcting PEF values for the known nonlinearity of mini-Wright PEF meters. Every 6 months, for a period of 2 yrs, PEF diary cards were filled in and BHR to histamine was tested in 104 patients with BHR and reversible airways obstruction, who started treatment with bronchodilators with (n=33) or without (n=71) inhaled corticosteroids. Within each subject, PEF indices and BHR were correlated longitudinally. The highest median correlation coefficients were obtained in the group of patients using inhaled corticosteroids. The PEF indices providing the best correlation with BHR were: mean PEF bronchodilator response (rho=-0.50) and within-day variation (% mean or % maximum) (with postbronchodilator values, rho=-0.50; without postbronchodilator values, rho=-0.40). Using PEF data corrected for the nonlinearity of the PEF meters did not result in higher correlation coefficients. Since current guidelines on asthma management recommend only bronchodilators on demand, the most useful peak expiratory flow index for reflecting bronchial hyperresponsiveness longitudinally is mean within-day peak expiratory flow variation (% mean or % maximum) (without postbronchodilator values). Since the correlation coefficients are not very strong, the authors suggest that peak expiratory flow measurements are not used as a proxy for bronchial hyperresponsiveness longitudinally but as a measurement in its own right. The use of corrections of peak expiratory flows for the nonlinearity of mini-Wright peak expiratory flow meters does not improve the correlation between peak expiratory flow and bronchial hyperresponsiveness.