The purpose of this study was to determine the relationship between fetal inflammatory and immune responses to oral pathogens and risk for preterm birth. Six hundred and forty umbilical cord blood specimens were prospectively collected. Cord serum levels of C-reactive protein, interleukin (IL)-1beta, IL-6, tumor necrosis factor (TNF)-alpha, Prostaglandin E2, and 8-isoprostane were determined by enzyme-linked immunosorbent assay and categorized as > median (high) versus < or = median (low). Presence of fetal immunoglobulin M (IgM) antibody against oral pathogens was determined by checkerboard immunoblot assay; detection of > or = 1 oral pathogen specific antibody was categorized as positive. Preterm birth was defined as spontaneous delivery at <35 weeks. Chi-square analysis was used to determine association between cord serum mediator or IgM category and preterm birth. Odds ratios (OR) for preterm birth were calculated, stratified by mediator and IgM category. Of 640 births, 48 (7.5%) delivered preterm. Preterm birth rates were higher if categorized as high versus low 8-isoprostane or TNF-alpha (23 vs 5%, P < .001 and 10 vs 4%, P < .01, respectively). Preterm birth rates were also higher if categorized as IgM positive versus negative (10.6 vs 5.8%, P = .04). The joint effects of fetal IgM seropositivity, detectable C-reactive protein, or high 8-isoprostane, PGE(2), or TNF-alpha resulted in significantly increased risk for preterm birth (adjusted OR [95% CI]: 6.0 [2.2-16.5], 4.3 [1.6-11.5], 4.1 [1.5-11.6], and 7.6 [2.3-20.8], respectively). Fetal exposure to oral pathogens evidenced by an IgM response is associated with preterm birth, and the risk for preterm birth is greatest among fetuses that also demonstrate an inflammatory response.