The authors conducted a cohort study within the Danish National Birth Cohort to determine whether coffee consumption during pregnancy is associated with late fetal death (spontaneous abortion and stillbirth). A total of 88,482 pregnant women recruited from March 1996 to November 2002 participated in a comprehensive interview on coffee consumption and potentially confounding factors in pregnancy. Information on pregnancy outcome was obtained from the National Hospital Discharge Register and medical records. The authors detected 1,102 fetal deaths. High levels of coffee consumption were associated with an increased risk of fetal death. Relative to nonconsumers of coffee, the adjusted hazard ratios for fetal death associated with coffee consumption of 1/2-3, 4-7, and > or =8 cups of coffee per day were 1.03 (95% confidence interval (CI): 0.89, 1.19), 1.33 (95% CI: 1.08, 1.63), and 1.59 (95% CI: 1.19, 2.13), respectively. Reverse causation due to unrecognized fetal demise may explain the association between coffee intake and risk of fetal death prior to 20 completed weeks' gestation but not the association with fetal loss following 20 completed weeks' gestation. Consumption of coffee during pregnancy was associated with a higher risk of fetal death, especially losses occurring after 20 completed weeks of gestation.