We compared growth of infants whose mothers either did or did not receive docosahexaenoic acid (DHA) supplements during pregnancy and lactation. At 21 weeks' gestation, 144 women were enrolled into a randomized, double-blind clinical trial receiving: (1) a basic supplement consisting of vitamins and minerals (BS), or (2) BS plus 4.5 g fructooligosaccharide (BSF), or (3) BSF plus fish oil DHA (200 mg) until the end of the third month of lactation. Infants' length, weight and head circumference were measured at birth and at 1, 3 and 21 months. A total of 51 mothers/infants were lost to follow-up by the third month and 24 at 21 months. The two groups not receiving DHA were combined into a control group. Analysis with mixed models adjusted for confounding factors showed a significant time dependent effect for the DHA group on the development of the body mass index (BMI) (P=0.037), and of weight (P=0.046), but no effect on the development of length (P=0.537), or of head circumference (P=0.267). At 21 months, weight of the DHA group was lower by -601 g (95% CI -171; -1030 g) and BMI was lower by -0.76 kg/m(2) (95% CI -0.07; -1.46) compared to controls. The results indicate that DHA taken by pregnant and lactating mothers may reduce BMI in late infancy.