To determine the effect of antimicrobial therapy on acute pulmonary exacerbations in cystic fibrosis, a randomized, double-blind trial of tobramycin and placebo was carried out. Clinical responses were satisfactory in all 11 children given tobramycin and in seven of 11 given placebo. Two patients in the placebo group died. No patient given placebo had improved results on pulmonary function studies, whereas improvement of 15% or more occurred in four of the six patients given tobramycin who could cooperate with the testing. Quantitative cultures of sputum showed a decrease of 1 logarithm or greater in Pseudomonas sp concentrations in six of seven patients in the tobramycin group and in two of eight in the placebo group. No difference in staphylococcal colonization was found. Several features indicate that children with severer disease were randomly assigned to the placebo group; nevertheless, the trend toward improved response in patients given tobramycin suggests that empirical therapy with antibiotics is beneficial for patients with acute pulmonary exacerbations in cystic fibrosis.