OBJECTIVE: The objective of this study was to examine whether the use of nicotine replacement therapy (NRT) during pregnancy increases the risk of stillbirth. DESIGN: Cohort study with prospective data. SETTING: Denmark 1996-2002. POPULATION: A total of 87,032 singleton pregnancies enrolled in the Danish National Birth Cohort for which information on NRT use as well as smoking was available. METHODS: Outcome of pregnancy was identified by register linkage, with <1% loss to follow up. We conducted Cox regression analyses to estimate the hazard ratio (HR) and 95% CI of stillbirth according to the use of NRT, type of NRT use and a combination of NRT use and smoking. MAIN OUTCOME MEASURES: Stillbirth, defined as delivery of a dead fetus after 20 completed weeks of gestation. RESULTS: A total of 495 pregnancies (5.7 in 1000 births) ended in stillbirth, 8 of which were among NRT users (4.2 in 1000 births). After adjustment for confounders, women who used NRT during pregnancy had a HR of 0.57 (95% CI 0.28-1.16) for stillbirth compared with those who did not use NRT. Smoking during pregnancy was associated with an increased risk of stillbirth (HR 1.46, 95% CI 1.17-1.82), while women who both smoked and used NRT had a HR of 0.83 (95% CI 0.34-2.00) compared with nonsmoking women who did not use NRT. CONCLUSION: Our study does not indicate that use of NRT during pregnancy increases the risk of stillbirth.