In a prospective, randomized, multicenter trial, the efficacy of penicillin plus ofloxacin was compared to that of amoxicillin-clavulanate plus erythromycin in the treatment of community-acquired pneumonia. One hundred seventeen hospitalized patients presenting with severe community-acquired pneumonia received either penicillin 3 x 10(6) U/6 h plus ofloxacin 200 mg twice daily (group A) or amoxicillin-clavulanate 1 g/6 h plus erythromycin 1 g/8 h (group B). Initial assessment included clinical examination, determination of simplified acute physiology score (SAPS), chest X-ray and evaluation of microbiological data obtained from blood, sputum and/or bronchoscopy. Follow-up was documented at 72 h and at 30 days. Both groups were comparable for age, sex, SAPS, chest X-ray, hypoxemia and microbiological data. The causative pathogen was identified in 54 cases (53%), Streptococcus pneumoniae being most frequent isolate (54.7%). All organisms cultured were susceptible to at least one of the antibiotics of each combination of the protocol, with the exception of two strains of Pseudomonas aeruginosa. A favorable outcome was observed in 76% of the patients, equally distributed between the two groups. After completion of therapy there were 12 clinical failures in each group (20.5%). Six patients in each group (10.3%) died of infection. Tolerance was similar for both regimens, apart from an increased rate of superficial thrombophlebitis in patients receiving intravenous erythromycin. The combination of penicillin with ofloxacin is as effective and as safe as a previously recommended regimen combining amoxicillin-clavulanate and erythromycin in treating patients with community-acquired pneumonia.