We examined the association between exposure to seafood intake during two periods of pregnancy on the one hand and risks of preterm delivery and postterm delivery on the other. In a prospective cohort of 8729 pregnant Danish women, we assessed frequency of fish meals during the first and second trimester of pregnancy by questionnaires completed around gestation weeks 16 and 30, respectively. When fish intake was based solely on intake reported for the early period of pregnancy, mean gestation length was shorter by 3.91 (95% CI: 2.24-5.58) days and odds of preterm delivery were increased 2.38 (1.23-4.61) times in those who never consumed fish (n = 308) vs. those who consumed both fish as main meal and fish in sandwiches at least once per week (n = 785). These measures were similar when fish intake was based solely on intake reported for mid-pregnancy. In the subgroup of women reporting same intake in the two trimesters, those who never consumed fish (n = 165) had 8.57 (5.46-11.7) days shorter mean gestation and 19.6 (2.32-165) times increased odds of preterm delivery, compared to high fish consumers (n = 127); odds of elective and postterm delivery were reduced by a factor 0.33 (0.11-1.02) and 0.34 (0.12-0.95), respectively, in zero fish consumers. All analyses were adjusted for potential confounding by factors such as maternal smoking, height, and prepregnant weight. We conclude that never consuming fish in the first two trimesters of pregnancy was an extremely strong risk factor for preterm delivery but was also associated with reduced risks of elective delivery and postterm delivery