The effect of anthropometric status on the efficacy of an oral supplement of iodised oil (1 ml Lipiodol Ultrafluide, 490 mg I; Laboratoire Guerbet, Aulnay-sous-Bois, France) was examined in 8–10-year-old schoolchildren (n 197) of Ntcheu, a severely I-deficient district of Malawi. The study was a controlled trial using the I concentration of casual urine samples to monitor the I status. The median urinary I concentration increased from 0·15 μmol/l at baseline (51·3 % of children < 0·16 μmol/l, 89·7 % < 0·40 μmol/l, 95·7 % 0·79 μmol/l) to 0·32 μmol/l at 40 weeks (29·1 % of the children < 0·16 μmol/l, 71·0 % < 0·40 μmol/l, 96·1 % < 0·79 μmol/l) while the total goitre prevalence fell from 63 % to 21 %. Variables of efficacy were estimated from a hyperbolic function describing the longitudinal pattern of urinary I excretion after the dose. The I retention and I elimination rate, and the periods of protection from mild (< 0·79 μmol/l) or moderate (< 0·40 μmol/l) I deficiency were obtained for groups of children with differing anthropometric status at baseline. Initial height-for-age and mid upper-arm circumference were not significantly related to efficacy. However, both the I retention and I elimination rate were reduced in children with lower initial weight-for-height. Children with lower skinfold thickness at baseline also had reduced I retention, which resulted in shorter protection periods from recurrent moderate and mild I deficiency. The efficacy of the oral iodised-oil supplement was not related to changes in anthropometric status during follow-up, nor was it related to the consumption of a food supplement of 1610 kJ immediately before the iodised-oil dose. Very low (< 0·16 μmol/l) urinary I concentration, and the presence of goitre at baseline were both associated with higher I retention and elimination rate. Children with goitre at baseline were found to have a prolonged duration of protection against recurrent moderate I deficiency. We conclude that in apparently healthy schoolchildren in I-deficient areas, general anthropometric status has a little influence on the efficacy of oral iodised oil for correcting I deficiency.