Recurrent and chronic lower airway infection with Pseudomonas aeruginosa (PA) is an important component of cystic fibrosis (CF) pulmonary disease. Different modes of treatment and control of CF patients have been introduced at the Copenhagen CF Centre over the past 20 years and have been associated with improved survival. Treatment consisted of: 1) elective antibiotics for 14 days every 3 months to patients with chronic PA infection (started in 1976), 2) cohort isolation to prevent cross-infection (patients with PA were separated from patients without PA, starting in 1981); and 3) early intensive treatment with inhaled colistin and oral ciprofloxacin from time of initial PA colonization (started in 1989). The aim of the present study was to evaluate the impact of each of these interventions on the changes in the epidemiology of PA. Based on monthly cultures of lower airway secretions in each CF patient seen during 1974-1995, significant changes in the incidence and prevalence of the PA infection were found. The monthly prevalence of chronic PA increased significantly (P < 0.0001) from below 40% before 1976 to above 60% in 1980, which was found to be due to cross-infection among the CF patients after introduction of elective antibiotic courses in 1976. To deal with this problem, cohort isolation was introduced in 1981, and since then the monthly point prevalence of chronic PA decreased slowly until 1989 (P < 0.0001), when early intensive treatment from initial PA colonization was introduced; this was associated with a further decrease in point prevalence to 45% in 1995 (P < 0.005). The annual incidence of chronic PA infection also decreased significantly (P < 0.01) from 16% to below 2% after introduction of cohort isolation and early intensive treatment from initial PA isolation. Furthermore, the time from acquisition of first PA to development of chronic PA infection increased significantly, from approximately 1 year to almost 4 years after introduction of cohort isolation (P < 0.0001). After introduction of early intensive treatment, the probability of still not having developed chronic PA infection 7 years after the first isolation of PA was above 80% (P < 0.0001). In conclusion, the introduction of cohort isolation and early intensive treatment following the initial isolation of PA resulted in a reduced incidence and prevalence of chronic PA infection. We are not aware of other studies showing a decreasing prevalence of chronic PA infection, as survival of CF patients has increased.