Riboflavin status by the erythrocyte glutathione reductase (NAD(P)H2:glutathione oxidoreductase EC1.6.4.2) test, and dietary riboflavin intake, were measured in 156 pregnant and lactating women in Keneba, a rural Gambian village, and in 59 pregnant and lactating women in Cambridge, England. The Keneba women were studied longitudinally, for up to 19 months. In Keneba, where the mean daily riboflavin intake was 0.5 mg, the mean erythrocyte glutathione reductase activation coefficient was 1.78 and there was a marked deterioration of biochemical status near parturition. In Cambridge, where the mean intake was 2.3 mg/day, the mean activation coefficient was 1.19. A vitamin fortified diet supplement, given to lactating women in Keneba for 8 months, raised their mean riboflavin intake for 0.5 to 1.5 mg/day and reduced the mean activation coefficient to 1.42. It is therefore likely that an intake even greater than the current United Kingdom or United States recommended daily amounts would be needed to achieve biochemical 'normality' in these women.