In a double-blind, double-dummy, multicentre study the efficacy and safety of dirithromycin 500 mg/day given orally once daily for 10-14 days were compared with those of erythromycin 1000 mg/day given orally four times daily for 10-14 days in patients with bacterial pneumonia. At 3-5 days post-therapy, 90 dirithromycin- and 83 erythromycin-treated patients were evaluable; the main reason for non-evaluability was failure to isolate the causative organism. Symptomatic responses were favourable in 94.5% of dirithromycin- and 100% of erythromycin-treated patients at post-therapy. At late post-therapy (2-3 weeks after completion of treatment), symptomatic responses were favourable in 98.7% of dirithromycin- and 94.8% of erythromycin-treated patients. At post-therapy, 63.3% of dirithromycin- and 50.6% of erythromycin-treated patients were unable to be evaluated bacteriologically, mainly due to Mycoplasma pneumoniae or Legionella pneumophila being assessed serologically. In the remaining evaluable patients treated with dirithromycin and erythromycin, bacteriological responses were favourable in 91.4% and 87.8%, respectively. At late post-therapy, favourable bacteriological responses occurred in 89.7% and 86.8%, respectively, of dirithromycin- and erythromycin-treated patients. Abdominal pain was the only treatment-emergent event to occur significantly more frequently in dirithromycin-treated patients.