ObjectiveTo determine whether a high concentration of serum ferritin during the third trimester is a marker of subclinical maternal infection and very preterm delivery and is associated with maternal nutritional status.MethodsA total of 1162 gravidas was followed prospectively from entry to prenatal care (15.0 ± 4.9 completed weeks' gestation) in Camden, New Jersey, between 1985 and 1995. Multiple logistic regression and analysis of covariance were used to examine the influence of serum ferritin on the outcomes of interest.ResultsHigh concentrations of serum ferritin (at or above the 90th percentile) at week 28, but not at entry to prenatal care, increased risk of preterm and very preterm delivery, but the risk changed if the concentration of ferritin declined from entry. If the concentration declined as expected, high ferritin concentration had no influence on outcome. If the concentration increased, then high ferritin concentration at week 28 was associated with very preterm delivery (adjusted odds ratio [AOR] 8.77; 95% confidence interval [CI] 3.90, 19.72), preterm delivery (AOR 3.81; 95% CI 1.93, 7.52), low birth weight (AOR 5.15; 95% CI 2.47, 10.72), clinical chorio-amnionitis (AOR 2.56; 95% CI 1.01, 6.52), and symptoms of “flu” as an index of unmeasured infection (AOR 6.02; 95% CI 1.16, 31.17). Factors associated with failure of the ferritin concentration to decline included iron deficiency anemia earlier in pregnancy (AOR 3.98; 95% CI 1.17, 8.98) and lower levels of serum and red cell folate.ConclusionHigh serum ferritin concentration in the third trimester, resulting from a failure of ferritin to decline, is associated with very preterm delivery and markers of maternal infection. Iron deficiency anemia and other indicators reflecting poor maternal nutritional status earlier in pregnancy underlie this relationship.