Diffuse panbronchiolitis (DPB) is characterized by chronic sinobronchial infection and diffuse bilateral micronodular pulmonary lesions consisting of inflammatory cells. Studies on disease etiology point to a genetic predisposition unique to Asians. Early therapy for DPB was largely symptomatic. The advent of macrolide antibiotics, including erythromycin, roxithromycin and clarithromycin, has strikingly changed disease prognosis. Low-dose, long-term macrolide therapy for DPB originated from detailed observations of response to therapy in a single patient. The bactericidal activity of macrolides, particularly erythromycin, is not a significant factor for their clinical efficacy in DPB. Firstly, irrespective of bacterial clearance, clinical improvement is observed in patients treated with erythromycin. Secondly, even in cases with bacterial superinfection with Pseudomonas aeruginosa resistant to macrolides, treatment has proved effective. Thirdly, the recommended dosage of macrolides produces peak levels in tissue that are below the minimum inhibitory concentrations for major pathogenic bacteria that colonize the airway. In the last two decades, the possible mechanism underlying the effectiveness of macrolide therapy has been extensively studied. The proposed mechanism of action includes inhibition of excessive mucus and water secretion from the airway epithelium, inhibition of neutrophil accumulation in the large airway, inhibition of lymphocyte and macrophage accumulation around the small airway, and modulation of bacterial virulence. The great success of macrolide therapy in diffuse panbronchiolitis may extend its application to the treatment of other chronic inflammatory disorders. If the anti-inflammatory activity of macrolides is independent of their bactericidal effect, new anti-inflammatory macrolides without antimicrobial activity should be developed to minimize emergence of macrolide-resistant micro-organisms.