Background Tobacco policy is increasingly focusing on the ‘tobacco endgame’ which commits to eradicating tobacco use (prevalence below 5%) within the next two decades. Strategies for achieving the endgame are likely to include addressing the supply of tobacco products, yet current evidence to support this approach is primarily cross-sectional.

Methods We use longitudinal smoking information from routine maternity records of all women who gave birth in Scotland between 2000 and 2015. We linked this data to the residential density of retailers selling tobacco products and the neighbourhood prevalence of smoking during pregnancy. In the analysis, individual mothers act as their own controls because we compare changes in their smoking behaviour between pregnancies to changes in exposure to tobacco retailing that arises from residential movement between pregnancies.

Results Adjusted ORs showed an increased risk of being a smoker associated with increases in exposure to retailer density (OR 1.67, 95% CI 1.27 to 2.20).

Conclusions The results provide the strongest evidence to date of an association between the neighbourhood availability of tobacco and smoking, and the first to do so among pregnant women. These findings provide supportive evidence for interventions targeting the supply of tobacco products in achieving the endgame.