Gambian women, who show marked seasonal fluctuations in energy balance, were offered a dietary supplement during pregnancy. This resulted in a net energy increment of 431 kcal/day. In the wet season, when the women were normally in marked negative energy balance due to food shortages and a high agricultural work load, the supplementation improved birth-weight by a mean of 224 g and reduced the incidence of low-birth-weight babies (less than 2.5 kg) from 28.2% to 4.7%. In the dry season, when the women were previously in positive energy balance despite an energy intake of only 60% of the recommended dietary allowance, the supplement had no beneficial effect on birth outcome. This threshold effect indicates the need to identify truly at-risk groups for prenatal supplementation programmes in developing countries.