We studied the clinical factors which have the greatest effect on bronchial hyperresponsiveness, with 37 atopic asthma patients (23 males and 14 females). They were from 13 to 59 years old. We measured the control value of the respiratory conductance (Grs.cont), the minimum dose of methacholine (bronchial sensitivity, Dmin), the linear slope of the Grs (= 1/Rrs) decreased (bronchial reactivity, SGrs) and SGrs/Grs.cont by the "Astograph" method (Chest 80, 600, 1981). Statistical analysis was performed by multiple regression analysis. Grs.cont, log Dmin, SGrs and SGrs/Grs.cont were regarded as objective variables. Age, onset age of asthma, period of disease, smoking history, family history of atopy, serum IgE, blood eosinophil counts, positive skin test counts, FEV1.0% and type of asthma attack (perennial or seasonal) were regarded as explanatory variables. The results were as follows: 1) The explanatory variable which showed the highest partial correlation coefficients, was the type of asthma attack in multiple regression analysis of Grs.cont and log Dmin (p < 0.05, p < 0.05). 2) Grs.cont of perennial asthma attack (0.247 +/- 0.064) was significantly lower than that of seasonal attack (0.318 +/- 0.097) (p < 0.02). Log Dmin of perennial asthma attack (-0.837 +/- 0.457) was significantly lower than that of seasonal attack (-0.254 +/- 0.429) (p < 0.005). Therefore in mild atopic asthma, the clinical factor which has the greatest effect on bronchial hyperresponsiveness is the type of asthma attack. We conclude that the existence of perennial asthma attacks is mostly related to increased bronchial hypersensitivity.