To assess mechanisms of bronchodilation and effectiveness of metered dose inhalers, it may be useful to determine sites of drug deposition in the lung. To establish suitable test aerosols, two brands of metered dose inhalers containing bronchodilator (Brethaire, Proventil) were radiolabeled with technetium ( 99mTc) and tested to determine if the distribution of radioisotope in the aerosol was representative of the distribution of agonist activity. Cascade impaction was used to determine the particle size distribution of the radioisotope and drug aerosols by assaying each state of the cascade using scintillation and HPLC techniques. Possible influences of the radiolabeling method and delivery techniques on the particle distribution were assessed by analyzing distributions from nonradiolabeled inhalers using HPLC. For these drugs, there was an excellent correlation between the distribution of radioactivity and the drug within the captured aerosol (Brethaire r = 0.994, Proventil r = 0.998, 20 - 200 consecutive puffs). Distributions were close to log-normal and differences in mass median aerodynamic diameter (MMAD) between the radioisotope and agonist activities were not significant (Brethaire, MMAD +/- sigma g, radiolabel vs drug = 4.7 +/- 2.1 mum, vs 4.4 +/- 1.7 mum, and Proventil, 2.5 +/- 2.1 mum, vs 2.4 +/- 2.0 mum. Non-labeled inhalers produced similar drug distributions (Brethaire, 4.2 +/- 1.8 mum, and Proventil 2.0 +/- 1.9 mum). Pausing between actuations resulted in slightly smaller distributions (Brethaire, 3.6 +/- 1.8 mum, Proventil 1.8 +/- 1.8 mum, 20 puffs-60 sec pauses) but the differences were not significant. In addition, to search for multimodal distributions and assess the accuracy of the MMAD measurement via our cascade impactor (Delron), we also measured the distribution of the mass of material within the aerosols using a weight-sensitive cascade (California Measurements). Using the latter device (1 puff), the mass distributions of both aerosols were similar to the values obtained from the puffs with pauses (Brethaire 3.8 +/- 2.3 mum, Proventil 1.4 +/- 2.2 mum). Multi-modal distributions were not found. By all assessments, the distributions were nearly log-normal with drug activity well described by the radiolabel.