In an effort to find a potential alternative treatment for scrub typhus, we evaluated the effectiveness of the standard drug doxycycline and the new macrolide azithromycin against a doxycycline-susceptible strain (Karp) and a doxycycline-resistant strain (AFSC-4) of Rickettsia tsutsugamushi. The antibiotics were tested in an in vitro assay system in which infected mouse fibroblast cells (L929) were incubated for 3 days in various concentrations of the drugs. Rickettsial growth was evaluated by direct visual counts of rickettsiae in Giemsastained cells or by flow cytometry. Initial tests were conducted at the concentration of each antibiotic considered to be the upper breakpoint for susceptibility (16 micrograms/ml for doxycycline and 8 micrograms/ml for azithromycin). Growth of both Karp and AFSC-4 was strongly inhibited with both antibiotics, as measured by visual counts, although the percentage of cells infected with AFSC-4 in the presence of doxycycline was three times greater than the percentage of cells infected with Karp but was only 60% as great as the percentage of cells infected with Karp in the presence of azithromycin. Flow cytometry confirmed that rickettsial growth occurred in the absence of antibiotics, but it failed to detect it in the presence of high concentrations of either drug. Visual counts of rickettsial growth at lower concentrations of the antibiotics (0.25 to 0.0078 microgram/ml) showed that the Karp strain was 16 times more susceptible that the AFSC-4 strain to doxycycline. Azithromycin was much more effective than doxycycline against AFSC-4, inhibiting rickettsial growth at 0.0156 microgram/ml to levels below that achieved by 0.25 microgram of doxycycline per ml. Azithromycin was also more effective than doxycycline against the Karp strain, causing greater reductions in the number of rickettsiae per cell at lower concentrations. If in vivo testing confirms the in vitro effectiveness of azithromycin, it may prove to be the drug of choice for the treatment of scrub typhus in children and pregnant women, who should not take doxycycline, and in patients with refractory disease from locations where doxycycline-resistant strains of R. tsutsugamushi have been found. When tested in an in vitro assay system, azithromycin was more effective than doxycycline against doxycycline-susceptible and -resistant strains of R. tsutsugamushi.