Floating fog occurs every summer in Kushiro City in Japan, and the annual average of fog water pH in the past 4 years has been under 5.0. We previously reported that epidemiologically fog was the most important positive factor contributing to increased hospital visits of asthmatic patients compared with other meteorological values and air pollutants. This study aimed to investigate the mechanism of the effects of naturally-occurring acid fog on asthmatic patients. We compared pulmonary functions and inflammatory mediators in induced sputum between the foggy (July 1995) and the non-foggy (May 1996) season, and assessed airway responsiveness to hypo-osmolar aerosol. Forty-four out of 118 asthmatic patients of Kushiro City residents participated, pulmonary function tests were completed in 36 patients, and sputum data were available in 26 patients in both seasons. Percent forced expiratory volume in 1 sec (FEV1) was significantly (P< 0.05) decreased, and % peak expiratory flow rate (PEFR) had a trend to decrease in the foggy season more than in the non-foggy, and sputum eosinophil cationic protein (ECP) and interleukin (IL)-8 were higher in the foggy season but not significantly. A moderate inverse correlation was revealed between sputum ECP and %PEFR in the foggy season (r= -0.55, P<0.005). Subjects were divided into two groups according to the best PEFR; one had >10% lower PEFR levels in the foggy season than in the non-foggy season (Group A, n = 7), the remainder did not (Group B, n = 19). In group A, sputum ECP was significantly increased (P< 0.01) in the foggy season, but there were no changes in IL-8 and prostaglandin D2. Ultrasonic nebulized distilled water provocation test revealed no differences between group A and B. These results suggested that eosinophilic inflammation rather than hypo-osmolar effect of fog might contribute to respiratory deterioration by inhalation of naturally-occurring acid fog.