Many pregnant women, in the world, drink caffeine-containing beverages. Maternal caffeine consumption during pregnancy may have adverse effects on foetus but results are conflicting. Our goals were to estimate the prevalence of caffeine use in a cohort of French pregnant women using maternal self-reports and to evaluate the association between caffeine consumption during pregnancy and delivery and new-born characteristics. All pregnant women who gave birth in a large French urban area during a limited period of time were included (in total 724 mothers were included). Coffee, tea or cola consumption as well as pregnancy and neonate characteristics were analysed. The mean consumption of caffeine per day slightly decreased from the first to the third trimester of pregnancy: 587 caffeine users, with a consumption of caffeine of 59.2 ± 61.5 mg/day during the first trimester as compared to 577 consumers (54.3 ± 55.4 mg/day) during the third trimester, respectively. A significant decrease of neonates’ birth length was observed when mothers were using at least 100 mg/day (or two cups) of caffeine during the second and third trimesters but this difference was no longer significant after adjustment on potential confounding factors such as tobacco use. The potential existence of other confounders (e.g. poorer dietary habits or other lifestyle variables) that might also be associated with reduced birth length, may not be excluded. Caffeine use during pregnancy was associated with reduced birth length but this effect was no longer significant after adjustment on potential confounding variables.