OBJECTIVE: To better understand the consequences of branded macrolide therapy, we assessed rates of re-treatment following the use of azithromycin or clarithromycin among outpatients diagnosed with common respiratory-tract infections in the California Medicaid ("Medi-Cal") program. METHODS: We employed a retrospective cohort design using administrative claims data for a 20% random sample of California Medicaid recipients. We selected outpatients diagnosed with common upper or lower respiratory-tract infections, including acute otitis media, acute pharyngitis, acute sinusitis, chronic bronchitis, or community-acquired pneumonia during an office visit between February 1, 1997 and May 31, 2002. Patients were prescribed either azithromycin or clarithromycin as treatment following diagnosis. The main study outcome was the filling of another prescription for an antibiotic (i.e., "retreatment") within 30 days following receipt of branded macrolide therapy. For each infection type, logistic regression was used to estimate the rate of retreatment for the two initial macrolides, adjusting for age. RESULTS: We identified 6196 patients diagnosed with acute otitis media, 3794 with community-acquired pneumonia, 2716 diagnosed with pharyngitis, 1354 with chronic bronchitis, and 1,162 with sinusitis who received one of the two macrolides of interest. Sixteen (acute otitis media) to 20% (sinusitis) of patients failed treatment with their initial macrolide. Relative to clarithromycin, azithromycin was associated with a significantly lower retreatment rate when used in patients with acute otitis media (OR: 0.63; 95% CI: 0.50 to 0.80; P <0.0001) or chronic bronchitis (0.70; 0.50 to 0.98; P = 0.039). The likelihood of retreatment (azithromycin relative to clarithromycin) did not differ for patients with pharyngitis (0.96; 0.74 to 1.26; P = 0.783), sinusitis (1.03; 0.75 to 1.42; P = 0.842), or community-acquired pneumonia (0.93; 0.77 to 1.13; P = 0.450). CONCLUSIONS: Compared to clarithromycin, azithromycin is associated with a lower retreatment rate when used for acute otitis media or chronic bronchitis in this Medicaid population.