The effect of two levels of energy supplementation in the last trimester of pregnancy on birthweight was tested in a controlled randomized trial in three villages in Madura, East Java. The high and low energy supplements provided 1.95 MJ (465 kcal) and 218 kJ (52 kcal) per day respectively. In the baseline period the home diet provided on average 6.28 MJ (1500 kcal) (SD 2.1 MJ (499 kcal] and 41 g (SD 13 g) of protein. The mean birthweight was 2835 g and the rate of low birthweight 12.2%. In the experimental period the home diet was better. The average intake ranged from 6.45 to 7.19 MJ (1541-1717 kcal) and 41.4-44.2 g per day, depending on the degree of compliance. Mean birthweight increased by 100 g and the rate of low birthweight dropped to 9.5%. There was no difference between the high and low energy supplemented group as a whole, probably due to the masking effect of the better home diet in the experimental period. It is likely that a positive effect of energy supplementation on birthweight was restricted to the group of pregnant women with the lowest home dietary intake and/or a low prepregnant weight. In this community targeting of supplementation to lean seasons and/or to women with a low prepregnant weight may be cost-effective.