Objective Preterm birth is a major cause of childhood mortality and morbidity, but the aetiology of preterm delivery remains incompletely understood. We sought to examine whether cumulative exposure to neighbourhood socioeconomic disadvantage is associated with preterm delivery risk.

Methods The association between neighbourhood socioeconomic disadvantage over the 20 years preceding delivery and risk of preterm delivery was assessed in a population-based cohort of 11 979 deliveries in Southwest Finland in 2008–2010. The statistical analyses were adjusted for a large number of potential individual-level confounding or mediating factors obtained from the national registry on parturients, deliveries and births.

Results Altogether 615/11 979 deliveries (5.1%) occurred before 37 weeks of gestation. The incidence of preterm delivery was 6.2% (95% CI 5.0% to 7.7%) in the women with the highest cumulative exposure to neighbourhood disadvantage as compared with 3.6% (95% CI 2.8% to 4.5%) in women who had lived in the most affluent neighbourhoods adjusted for the covariates; OR 1.74 (95% CI 1.26 to 2.40). Smoking during pregnancy, prepregnancy body mass index, gestational diabetes, pre-eclampsia and other medical problems during pregnancy explained 27.4% of this association. Shorter exposure to neighbourhood disadvantage was associated with lower excess risk of preterm births.

Conclusions Women with long-term exposure to neighbourhood socioeconomic disadvantage are at increased risk for preterm delivery in a dose-dependent fashion. Improving deprived neighbourhoods may offer means to reduce the risk of preterm birth and, consequently, the intergenerational transfer of health inequality.