Pulmonary infection in cystic fibrosis (CF) is primarily a purulent tracheobronchitis. Antibiotics are available that are active in vitro against bacteria isolated from sputum from patients with CF. Despite efficacious antibiotic concentrations in serum, however, the results of treatment are frequently suboptimal. A widely accepted explanation for this limited efficacy is poor penetration of orally or intravenously administered antibiotics into respiratory secretions. The bioactivity of antibiotics in respiratory secretions is not identical to that found in vitro. Laboratory conditions are standardized and selected to approximate serum. Deviations from these conditions can markedly influence the results. Differences in composition between sputum and laboratory culture media, as well as variation in growth and metabolism of the pathogen in respiratory secretions, must be considered when predicting in vivo activity in sputum. Thus, when defining criteria for antibiotic susceptibility or resistance in the treatment of pulmonary infection in patients with CF, the concentrations achievable in bronchial secretions as well as the bioactivity in this environment should be considered.