The aim of this paper was to estimate the effect of maternal and paternal smoking on foetal death (miscarriage and stillbirth) and to estimate potential interactions with physical exercise and pre-pregnancy body mass index. We selected 87,930 pregnancies from the population-based Danish National Birth Cohort. Information about lifestyle, occupational, medical and obstetric factors was obtained from a telephone interview and data on pregnancy outcomes came from the Danish population based registries. Cox regression was used to estimate the hazard ratios (adjusted for potential confounders) for predominantly late foetal death (miscarriage and stillbirth). An interaction contrast ratio was used to assess potential effect measure modification of smoking by physical exercise and body mass index. The adjusted hazard ratio of foetal death was 1.22 (95 % CI 1.02-1.46) for couples where both parents smoked compared to non-smoking parents (miscarriage: 1.18, 95 % CI 0.96-1.44; stillbirth: 1.32, 95 % CI 0.93-1.89). On the additive scale, we detected a small positive interaction for stillbirth between smoking and body mass index (overweight women). In conclusion, smoking during pregnancy was associated with a slightly higher hazard ratio for foetal death if both parents smoked. This study suggests that smoking may increase the negative effect of a high BMI on foetal death, but results were not statistically significant for the interaction between smoking and physical exercise.