Objective: To compare the safety and efficacy of clarithromycin and amoxicillin/ clavulanic acid in patients with community-acquired pneumonia due to penicillin-resistant and/or macrolide-resistant Streptococcus pneumoniae, by selecting clinical investigators who practice in study populations from geographic areas in which a high incidence of resistant strains is reported by surveillance. Design and setting: Prospective, randomised, investigator-blinded, multicentre study conducted in 45 sites in primary-care and referral centre settings. Patients and interventions: 327 ambulatory patients diagnosed with radiographically confirmed community-acquired pneumonia administered clarithromycin 500mg immediate-release or amoxicillin/clavulanic acid 875mg/125mg twice daily for 7 days. Main outcome measures and results: Similarly high clinical cure rates were observed among evaluable patients in both treatment groups at the test-of-cure visit (28-35 days post-treatment): 92% (114/124) for clarithromycin and 91% (117/129) for amoxicillin/clavulanic acid. Of 85 S. pneumoniae strains isolated pretreatment, four (5%) were classified as resistant to macrolides (one mefA, two ermB, and one ermB + mefA) and eight (9%) had reduced susceptibility to penicillin. The overall eradication rate for pathogens isolated from bacteriologically and clinically evaluable patients was 91% for clarithromycin and 93% for amoxicillin/clavulanic acid, and 89% and 92%, respectively, for S. pneumoniae strains. The rates of resolution and/or improvement in clinical signs and symptoms and radiological improvement were similar with clarithromycin to those with amoxicillin/clavulanic acid, as was overall incidence of adverse events. Conclusion: A 7-day course of clarithromycin immediate-release was similar to amoxicillin/clavulanic acid based on high rates (>90%) of clinical cure, radiological improvement and pathogen eradication among ambulatory-care patients with community-acquired pneumonia. As the resistance rate at baseline was low, no conclusion could be made about clarithromycin's efficacy for infections caused by macrolide-resistant S. pneumoniae. Both treatments were well tolerated.