Background: Eosinophil infiltration of bronchial tissues and subsequent release of inflammatory mediators by them are the hallmarks of bronchial asthma but it has not yet been clarified whether anti-asthma drugs affect these cells directly. In this study, we investigated the direct effects of 8 clinically used anti-asthma drugs [salbutamol, salmeterol, theophylline, denbufylline, disodium cromoglycate (DSCG), azelastine, ketotifen and dexamethasone] on superoxide anions (O–2) and eosinophil peroxidase (EPO) release from human blood eosinophils in vitro. Methods: Highly purified eosinophils were stimulated for O–2 release with platelet-activating factor (PAF) or interleukin-5 (IL-5), while for EPO release complement fragment (C5a) or N-formyl-methionyl-leucyl-phenylalanine (FMLP) was employed. Generated products were assayed by standard techniques. Results: All the drugs, except ketotifen and dexamethasone, inhibited PAF-induced O–2 release in a dose-dependent manner. The IC50 values were 0.7, 5.8, 330, 3,500, 4,200 and 6,250 nM for DSCG, denbufylline, salmeterol, azelastine, salbutamol and theophylline, respectively. On IL-5-induced release, the effects were similar except that salbutamol completely failed to inhibit the release induced by this stimulus. In contrast, EPO release was generally poorly inhibited, especially when the release was induced by C5a. Only theophylline and azelastine (both at 10–4 M or more) were able to inhibit EPO release by both C5a and FMLP. Salbutamol and, to a lesser extent, salmeterol inhibited FMLP-, but not C5a-induced EPO release, while all the other drugs tested were inactive. Conclusions: The results show that some of the anti-asthma drugs, but not all, do exert direct effects on human blood eosinophils but these effects may be stimulus-dependent and by far more pronounced against O–2 release than against degranulation.