BACKGROUND: Doxycycline, azithromycin, and chloramphenicol have been used for the treatment of scrub typhus; however, the duration of treatment for this disease is still controversial. Most clinical studies on the treatment of scrub typhus focus on the effectiveness of antibiotics; few studies have focused on the duration of the treatment. There has been no in vitro evaluation of the optimal duration of treatment of scrub typhus. We performed an experiment to evaluate the regrowth of Orientia tsutsugamushi after exposure to doxycycline of various concentration and exposure time. MATERIALS AND METHODS: O. tsutsugamushi strain Boryong was inoculated into the ECV304 cell line. The infected cells were stained with FS15, a monoclonal antibody reacting against a linear epitope on the 56-kDa major outer membrane protein of O. tsutsugamushi; thereafter, the antimicrobial susceptibilities were measured by flow cytometry and expressed as a growth index (total mass of Orientia). Incubation in media containing 0.1 (the minimal inhibitory concentration of doxycycline against O. tsutsugamushi strain Boryong), 0.2, and 2 microg/mL doxycycline for 1, 3, 5, and 7 day was followed by incubation in antibiotic-free media for 5 day. Two inocula sizes were employed; low inoculum (45.18% of cells were infected) and high inoculum (87.21% of cells). RESULTS: The incubation with 0.1 microg/mL doxycycline resulted in O. tsutsugamushi regrowth irrespective of doxycycline treatment duration. When the concentration of doxycycline was increased to 0.2 or 2 microg/mL, regrowth was observed in short duration treatments (1-5 day at 0.2 microg/mL; 1 day at 2 microg/mL). Higher inoculum size of O. tsutsugamushi increased regrowth (1 day at the low inoculum; 1-5 day at the high inoculum). Conclusions: The regrowth of O. tsutsugamushi after doxycycline treatment depends on the doxycycline dose; the higher the dose, the lower the relapse rate. If this result is extrapolated to humans, higher dose of doxycycline might be preferred to lower the relapse rate. The inoculum effect in O. tsutsugamushi infection needs further investigations.