The therapeutic doses of beta-stimulants in the treatment of bronchial asthma show wide variation among the drugs. In the present study, we tried to construct a system for prediction of the optimum clinical dose of beta-stimulants as oral dosage forms and aerosols using the data from preclinical pharmacological studies and pharmacokinetic data in humans. Values of EC50 (drug concentration that gives the half-maximum bronchodilator effect) in the in vitro studies based on isolated trachea of guinea pigs were collected from the literature. The ratios of the EC50 value of each drug to that of isoproterenol were used as the indicator of in vitro pharmacological activity. There were significant correlations between the EC50 ratios and the oral dose or the maximum plasma concentration after single oral administration, but the correlation coefficients were relatively small (r < .9). On the other hand, a linear log-log relationship was observed between the EC50 ratio and the maximum plasma unbound concentration (Cu(max)) (slope = 0.907, r = .955, p < 0.001). In the case of aerosols, a good correlation was observed between the EC50 ratio and the dose (slope = 0.770, r = .900, p < 0.01). The predicted concentration of beta-stimulants in the lung after aerosol administration tended to be higher than the Cu(max) after oral administration, suggesting the contribution of nonspecific binding in the lung tissue. These findings indicate the possibility of predicting the appropriate dose of beta-stimulants based on the preclinical pharmacological data and the pharmacokinetic data in humans.