Clinical studies comparing the potency of inhaled corticosteroids require steep dose-response slopes (b) and minimal response variability (s), as statistical power is inversely related to the s/b ratio. To evaluate a new study model, we performed a randomized, crossover study of 12 adult asthmatics who required 800 to 2,000 microg of inhaled corticosteroids daily, and calculated s/b for 21 raw clinical outcomes and 36 mathematically derived variables based on these raw outcomes. Each of two 21-d treatment periods was preceded by 4 to 7 d of oral prednisone to maximize asthma control and minimize carry-over of previous inhaled treatment. Treatments were 100 and 800 micron/d of an HFA-134a beclomethasone dipropionate formulation. Assessments included daily home spirometry, histamine challenge, inhaled albuterol use, and asthma symptom scores. Efficacy variables with the greatest power (lowest s/b values) were A.M.FEF25-75, A.M.FEV1, and A.M.PEF, (s/b = 0.46, 0.48, and 0.59). Carry-over between treatment periods was not significant. Crossover study sample size calculations using these ratios yielded samples of 23, 25, and 37 patients, respectively. Otherwise identical parallel studies would require sample sizes of 657, 1,438, and 2,261 patients. These results support the use of a crossover asthma stability model after a short course of oral prednisone as a clinical study model for comparing topical potency of inhaled corticosteroids.