Fifteen asthmatic patients were prospectively selected. If they had more than 450 eosinophils in peripheral blood on admission to hospital treatment and if before BAL had been performed they were not treated by corticosteroids. No associated disease was found which could cause the increased number of eosinophils in peripheral blood except bronchial asthma. BAL was performed when the subjects tested had no dyspnea, usually 1-3 days after the hospital admission. In contrast to blood eosinophilia, increase of eosinophils number in BAL was found to correlate to impairment of clinically measured parameters such as: ventilation level of blood gases and duration of hospitalisation. On the contrary there was no correlation between the level of blood eosinophilia and gas disturbance and duration of hospitalisation. Eosinophilic proteins (MBP, ECP, EPX, EPO) are not the only that harmfully effect the lung cells in the asthmatic attack but the fact that their level in lavage fluid correlates to eosinophilia degree in BAL is evident so the correlation between the number of eosinophils in BAL and impairment of clinically measured parameters is clear.