To determine the prevalence of nontuberculous mycobacterial infection in patients with cystic fibrosis. We performed a prospective study in which patients with cystic fibrosis were followed for 2 years; the patients were recruited from specialized units and were all over 6 years old. Sputum samples collected every 6 months were stained with auramine-rhodamine and cultures were prepared with a liquid and a solid medium. When stains or cultures were positive for nontuberculous mycobacteria, 1 or 2 additional sputum samples were obtained from the patients, who were monitored closely to assess the need for specific treatment. We assessed the following clinical variables: age, sex, presence of pancreatic insufficiency, use of aerosol antibiotic therapy, and long-term azithromycin and inhaled or oral corticosteroid therapies. A total of 220 patients (119 women) with a mean age of 22.62 years (range, 6-74 years) were enrolled; of these 23.6% were receiving azithromycin. We prepared 1303 sputum samples for mycobacterial growth (range per patient, 4-68 samples); 65 samples from a total of 17 patients (7.72%) were positive: 17 by auramine-rhodamine staining and 48 by culture. Eighty-eight culture samples were contaminated and Mycobacterium tuberculosis was not isolated in any of the cases. The mycobacteria isolated were M avium complex (n=10), M abscessus (n=6), and M fortuitum (n=1). Two or more positive cultures were obtained in 9 patients, 5 of whom experienced clinical deterioration and were prescribed specific treatment. No significant differences in clinical variables were found between patients with nontuberculous mycobacteria and those without. The prevalence of nontuberculous mycobacterial infection in patients with cystic fibrosis was not very high (7.72%), perhaps because azithromycin interfered with the growth of these bacteria. Patients with repeat isolations of mycobacteria should be monitored closely.