Summary. A prospective cohort study with a 1-year follow-up of 156 neonates was carried out specifically designed to test the hypothesis that there is a positive relationship between iron deficiency during pregnancy and the development of the same disease in newborn infants. Exposure was defined as being born of a mother with ferropenic anaemia at delivery, and cases as the infants who developed iron deficiency during their first year of life. A statistically significant positive association was detected with an odds ratio of 6.57 (95% confidence limits 1.81–25.97). A stratified analysis was also performed to control the effect of potential confounders such as socio-economic variables, feeding practices and other factors linked with the iron status of infants. This second analytical procedure showed no alteration in the association detected in the simple analysis but that there was a statistically significant strong interaction between the quantity of cow's milk intake and the ferropenic status of the mother. These results show a relationship between iron deficiency of the mother at delivery and the development of iron deficiency in the infants. These new findings could be important in the development of new prevention programmes applied to pregnant women.