Inhaled antiasthmatic steroids have been assumed and yet never proved to exert their antiasthmatic effect by topical action in the airways. We tested the hypothesis that the efficacy of inhaled budesonide (BUD) might be due instead to its systemic activity after absorption. We compared inhaled and oral BUD with doses selected to ensure higher peak plasma levels and a greater area under the plasma concentration curve with the oral treatment. After pretreatment with beclomethasone to maximize asthma control, 47 adults with asthma were randomized to receive 0.4 mg of inhaled BUD per day (n = 16) or 1.4 mg of oral BUD per day (n = 15), or placebo (n = 16) in double-blind fashion and then followed weekly until asthma relapsed or for 8 weeks if no relapse occurred. "Relapse" was defined as a drop in the mean peak expiratory flow rate greater than 2 SEM below the mean during the baseline week before switching to the test drugs. The time to relapse was the primary outcome variable. Time to relapse was longer with inhaled than with oral BUD (medians, 22 versus 7.9 days; p = 0.003) or placebo (medians, 22 versus 9 days; p = 0.004). Oral BUD and placebo did not differ (p = 0.41). The morning serum cortisol levels remained normal during all three treatments. Thus, at conventional dosage the antiasthmatic effect of inhaled BUD may be fully explained by a local intrapulmonary action.