Eighty-one pregnant or lactating women living in a village in The Gambia were allocated to treatment groups to receive a daily placebo, riboflavin, ferrous sulphate or ferrous sulphate plus riboflavin. At the beginning of the study and at 3 and 6 weeks thereafter women were examined clinically and blood samples collected for haematological and biochemical measurements. Pregnant women showed a wide range of response to supplement but among the lactating women an interesting pattern emerged: whereas the haematological status of the placebo group declined over 6 weeks, in the group receiving iron supplements this deterioration was reduced although not significantly. Riboflavin given in addition to iron resulted in a significant increase in circulating plasma iron and in iron stores, relative to the placebo.