Babies born small-for-gestational age (SGA) have an increased risk of mortality, morbidity and adverse functional consequences. Studies suggest that pre-pregnancy maternal diet may influence newborns’ size. This study aimed to determine whether maternal pre-pregnancy dietary patterns (DP) are associated with delivering SGA newborns in the ProcriAr Cohort Study, Sao Paulo-Brazil. Pre-pregnancy DP of 299 women were investigated using factor analysis with principal component’s estimation, based on intake reported on a validated 110-item FFQ. Newborns were classified as SGA if their weight and/or length, adjusted by gestational age and sex, were below the 10th percentile of the INTERGROWTH-21st standards. Multivariate Poisson regression modelling with robust error variance was performed to examine associations between the different DP (in quintiles) and SGA. In a model adjusted by maternal sociodemographic and health behaviours, women who scored in the highest quintile of the DP ‘Snacks, sandwiches, sweets and soft drinks’ (in relation to the women who scored in the lowest quintile) were significantly more likely to deliver SGA babies (relative risk 1·92; 95 % CI 1·08, 3·39). This study verified that women’s pre-pregnancy dietary behaviour characterised by an energy-dense nutrient-poor food intake was a risk factor for delivering SGA newborns. Investments in education and improved access to healthful food and nutritional information before pregnancy should be prioritised due to their potential positive impact on child health. However, further studies are warranted to identify specific metabolic pathways that may be underlying these associations.