OBJECTIVES: Maternal dental periapical infections are associated with preterm birth and intrauterine growth restriction. This study investigates whether the association is mediated through bacterial spread from periapical lesions to placenta (direct pathway) or systemic inflammatory reaction (indirect pathway). MATERIALS AND METHODS: We compared birth outcomes in Malawian mothers with and without periapical infection. As markers of a direct pathway, we identified placental bacteria using a 16S rDNA approach and assessed histological evidence of inflammation in the placenta and amniotic membranes. We measured C-reactive protein, alpha-1-acid glycoprotein, and salivary cortisol as markers of an indirect pathway. We used regression models to associate the predictor variables with duration of pregnancy and newborn size. RESULTS: Of 1,024 women, 23.5% had periapical infection. There was no association of periapical infection with either bacterial DNA or histological inflammation in placenta or membranes. Periapical infection was associated with C-reactive protein, alpha-1-acid glycoprotein, and cortisol concentrations in a dose-dependent manner at 36 weeks. Addition of alpha-1-acid glycoprotein or cortisol concentration into regression models attenuated the association between periapical infection and pregnancy outcomes. CONCLUSION: There was no evidence of direct spread of periapical bacteria to the placenta. Periapical infections and adverse pregnancy outcomes are in part mediated through systemic inflammation.