Two hundred and forty-two patients over 16 years of age with community-acquired lower respiratory tract infection (LRTI), matched for age and sex, were randomised to receive either roxithromycin 150 mg b.i.d. or amoxycillin 500 mg/clavulanic acid 125 mg t.i.d. for 7 days, with a further 7 days if insufficient response was seen. Clinical efficacy at 7 days was 69% for roxithromycin and 56% for amoxycillin/clavulanic acid (p = 0.05) and at study end it was 91% for both antibiotics. There were fewer second treatment courses in the roxithromycin group (26% vs. 38%, p = 0.04) and a shorter treatment duration (8.29 days vs. 9.34 days, p > 0.05). Twelve patients (9.8%) treated with roxithromycin and 19 (17.1%) treated with amoxycillin/clavulanic acid had adverse effects possibly, or probably, related to the antibiotic. Roxithromycin appears to be a more appropriate choice than amoxycillin/clavulanic acid for the treatment of LRTI in the community given its more appropriate in vitro spectrum, efficacy against most common and atypical pathogens, greater cost-effectiveness, more convenient dosage regimen (b.i.d.), and superior tolerability profile.