In a recent investigation we found that hospitalized patients with cystic fibrosis who received 0.5 cc of 0.5% albuterol nebulizer solution TID significantly increased their pulmonary function across the day, but fell back to baseline overnight. To determine whether this fall could be prevented by the long-acting beta-2 agonist salmeterol at both standard (2 puffs: 42 mcg BID) and high (4 puffs, 84 mcg BID) doses, we evaluated the effects of salmeterol vs. albuterol (2 puffs, 180 mcg QID, and 4 puffs, 360 mcg BID) in a placebo-controlled three-way random crossover, double-blind trial. Eighteen patients in the low-dose group and 10 of the same 18 patients in the high-dose group completed the 3 consecutive days of testing and received either salmeterol, albuterol, or placebo with each of four chest physiotherapy sessions given at 7 AM, 11 AM, 3 PM, and 7 PM. At standard doses (2 puffs), the mean percent changes in FEV1 pre- to post-7 AM therapy for salmeterol (5.5%) and albuterol (9.9%) were significantly greater than with placebo (-1.2%) (P < 0.05 and 0.01, respectively). The mean percent changes in FEV1 from morning baseline with salmeterol were also significantly greater than placebo before 3 PM (12.1% vs. 5.4%, P < 0.01), and neither albuterol nor salmeterol were significantly greater than placebo after 3 PM. At standard doses there was a significant carryover effect with salmeterol to the next morning for the FEV1 (7.3%) when compared to placebo (1.5%) and albuterol (-0.7%) (P < 0.05 and 0.05, respectively). At high doses (4 puffs), the mean percent change in FEV1 with pre- to post-7 AM therapy increased to 22.7% and remained significantly greater than with placebo until pretherapy at 7 PM. The carryover effect the next morning was 14.7%. Salmeterol at 4 puffs compared favorably to albuterol nebulizer therapy given TID in both the incidence of responders for the FEV1 (70% vs. 71%) and the mean changes after therapy at 7 AM (22.7% vs. 14.9%), and provided greater carryover effects to the next morning (14.7% vs. -0.7%), thus preventing the fall in pulmonary function back to baseline overnight. We recommend the use of high-dose salmeterol in hospitalized patients with FVC values of 40% of predicted or greater, starting with 2 and increasing to 4 puffs BID as tolerated.