The disease of asthma remains poorly understood. Hence, considerable latitude has been afforded to pharmacologists concerned to identify novel compounds for development as anti-asthma drugs. Compounds can be selected (1) on the basis of known efficacy as bronchodilator (beta-adrenoceptor agonists, parasympatholytics and xanthines) and prophylactic (cromoglycate and ketotifen) or anti-inflammatory (glucocorticosteroids) actions; (2) by reference to current theories of asthma pathogenesis, e.g., inhibition of the formation and action of autoacoids (leukotrienes, platelet-activating factor) or cytokines (GM-CSF, IL-5), or (3) by reference to hitherto unexplored physiological processes (K+ channel activation, PDE isoenzyme inhibition, selective immunosuppressive actions). There is a need to test whether the rationale for drug selection is justified. This can only be resolved by study of asthma patients, since normal subjects may respond inappropriately. Thus, normal subjects do not exhibit airway obstruction or airway hyperreactivity and hence do not respond to bronchodilator substances. Similarly, normal subjects respond to platelet-activating factor with a cellular infiltrate in which neutrophils predominate in contrast to atopic asthmatics who exhibit an eosinophilia. It follows that compound discovery and evaluation must be an iterative procedure in which animal studies are supplemented by clinical studies and vice versa. This is most readily achieved by establishing clinical research that is a complementary to preclinical research.