We investigated the in vitro and in vivo activities of macrolides against Mycoplasma pneumoniae. In vitro MICs of azithromycin, erythromycin, clarithromycin, and roxithromycin were determined. Azithromycin was the most potent antimicrobial agent tested in vitro. Its MIC for 90%o of the strains was 0.00024 ,ug/ml. MICs for 90%o of the strains of erythromycin, clarithromycin, and roxithromycin were 0.0156, 0.0078, and 0.03125 ,ug/mI, respectively. In vivo activities were assessed in a pulmonary infection model with Syrian golden hamsters. We evaluated the in vivo effects on reduction of viable M. pneumoniae cell counts and on reduction of microscopic and macroscopic histopathologies for azithromycin, erythromycin, and clarithromycin given at 10 mg/kg once daily for 1 and 3 days and given at 15 mg/kg twice daily for 2.5 and 5 days. Mycoplasma pneumoniae is a major cause of pneumonia and accounts for as many as 20% of the total number of cases of pneumonia (6). The recommended therapy is erythromycin, which is efficacious in reducing the duration of symptoms (19). Recently, it has been suggested that new macrolides (8, 10), such as azithromycin, clarithromycin, and roxithromycin, may be used more frequently for respiratory tract infections because of their improved potencies and pharmacokinetic parameters compared with those of erythromycin (10). Azithromycin is a new azalide antibiotic (7, 15) that differs from erythromycin by a methyl-substituted nitrogen at position 9a within the macrocyclic ring. This modification has resulted in improved potency against gram-negative bacteria. Moreover, azithromycin proved to be superior to erythromycin in a mouse pneumococcal model because of these improved pharmacokinetic parameters (1).In this study we evaluated the in vitro antimycoplasmal activities of azithromycin, erythromycin, clarithromycin, and roxithromycin. We also tested the in vivo antimycoplasmal efficacies of azithromycin, erythromycin, and clarithromycin in courses of 10 mg/kg once daily for 1 or 3 days or 15 mg/kg twice daily for 2.5 or 5 days by using an experimental pulmonary infection model with Syrian golden hamsters (2). Organisms. Forty clinically isolated strains of M. pneumoniae obtained from 1988 to 1992 at Nagasaki University Hospital and three standard strains, FH, Mac, and M129, which were supplied by M. F. Barile (Food and Drug Administration, Bethesda, Md.), were used for in vitro tests, and the virulent strain M129 at passage level 12 was used for experimental mycoplasmal pneumonia in Syrian golden hamsters.MIC determinations. MICs were determined by using a modified broth microdilution method (12, 22, 23) with modified Chanock broth medium (3), which consisted of seven parts PPLO broth without crystal violet (Difco Laboratories, Detroit, Mich.), 2 parts uninactivated horse serum, 1 part 25% fresh yeast extract, 1% glucose, and 0.002% phenol red adjusted to pH 7.8 with 1 N sodium hydroxide. Inocula were 10 CFU/ml. The plates, sealed with plate sealers, were incubated at 37°C in air. In each case...