OBJECTIVE: The purpose of this study was to assess the association of presumptive Trichomonas vaginalis treatment during pregnancy and birth outcomes. STUDY DESIGN: A community-randomized trial of presumptive sexually transmitted disease treatment during pregnancy was conducted between 1994 and 1999 in Rakai district, Uganda. A subanalysis of a trial of presumptive therapy with azithromycin, cefixime, and metronidazole assessed Trichomonas vaginalis treatment in pregnant women. RESULTS: Children of 94 women with Trichomonas who were treated had increased low birth weight (relative risk, 2.49; 95% CI, 1.12-5.50), preterm birth rate (relative risk, 1.28; 95% CI, 0.81-2.02), and 2-year mortality rate (relative risk, 1.58; 95% CI, 0.99-2.52), compared with children of 112 women with Trichomonas who were not treated. CONCLUSION: Treatment of Trichomonas vaginalis during pregnancy may be deleterious, and we infer that this may be due to metronidazole. This is consistent with a National Institute for Child Health and Human Development trial that found an excess of preterm births in children of women with Trichomonas vaginalis infection who were treated with metronidazole.