CONTEXT: Treatment recommendations assume that repeated mass antibiotic distributions can control, but not eradicate or even locally eliminate, the ocular strains of chlamydia that cause trachoma. Elimination may be an important end point because of concern that infection will return to communities that have lost immunity to chlamydia after antibiotics are discontinued. OBJECTIVE: To determine whether biannual treatment can eliminate ocular chlamydial infection from preschool children and to compare results with the World Health Organization-recommended annual treatment. DESIGN, SETTING, AND PARTICIPANTS: A cluster-randomized clinical trial of biannual vs annual mass azithromycin administrations to all residents of 16 rural villages in the Gurage Zone, Ethiopia, from March 2003 to April 2005. INTERVENTIONS: At scheduled treatments, all individuals aged 1 year or older were offered a single dose of oral azithromycin either annually or biannually. MAIN OUTCOME MEASURE: Village prevalence of ocular chlamydial infection and presence of elimination at 24 months in preschool children determined by polymerase chain reaction, correcting for baseline prevalence. Antibiotic treatments were performed after sample collections. RESULTS: Overall, 14,897 of 16,403 eligible individuals (90.8%) received their scheduled treatment. In the villages in which residents were treated annually, the prevalence of infection in preschool children was reduced from a mean of 42.6% (range, 14.7%-56.4%) to 6.8% (range, 0.0%-22.0%) at 24 months. In the villages in which residents were treated biannually, infection was reduced from 31.6% pretreatment (range, 6.1%-48.6%) to 0.9% (range, 0.0%-4.8%) at 24 months. Biannual treatment was associated with a lower prevalence at 24 months (P = .03, adjusting for baseline prevalence). At 24 months, no infection could be identified in 6 of 8 of those treated biannually and in 1 of 8 of those treated annually (P = .049, adjusting for baseline prevalence). CONCLUSION: Local elimination of ocular chlamydial infection appears feasible even in the most severely affected areas, although it may require biannual mass antibiotic distributions at a high coverage level. TRIAL REGISTRATION: clinicaltrials.gov Identifier: NCT00221364.