Objective: To examine and compare the effects of pregnancy on the thyroid hormone homeostasis in two different populations with variable iodine supply. Design: A longitudinal prospective cohort study throughout pregnancy involving Swedish and Sudanese pregnant women. Setting: The subjects were enrolled consecutively during their antenatal follow-up at health centres at Nyby in Uppsala, Sweden and Omdurman in Sudan. Subjects: Fifty-one apparently healthy women from Uppsala, Sweden and twenty-eight pregnant women from Omdurman, Sudan were recruited during pregnancy. The mean age and weight of the Swedish women at the beginning of pregnancy were 29.9 AE 5.4 y and 66.3 AE 12.9 kg respectively. The corresponding ®gures for the Sudanese women were 28.0 AE 4.9 y and 64.8 AE 9.4 kg respectively. Methods: Blood samples were drawn on four occasions from the Swedish group at 11±13, 24, 32, and 38 weeks of pregnancy, and on three occasions from the Sudanese group at 10±12, 20±24, and 36±39 weeks. Twenty-four hour urine samples were collected from the same subjects and on the same occasions as blood sampling. The urine samples were kept in a refrigerator until the volumes were measured, after which 20 mL aliquots were taken and kept frozen until analysed. No signi®cant changes in daily urinary iodine loss were observed in the two groups with progression of pregnancy. However, in the Swedish women the mean free T4 concentration fell from 11.81 pmol/l at the beginning of pregnancy to 8.82 pmol/l and the mean TSH rose from 1.11±1.95 mU/l between the beginning and end of pregnancy. Such changes were not detected among the Sudanese women, who had signi®cantly lower mean TSH values than the Swedish women in weeks 36±39 of pregnancy (P`0.02), and signi®cantly higher FT4 values than the Swedish women both in weeks 20±24 and in weeks 36±39 (P`0.005 and P`0.001) respectively. Conclusions: The study suggests that determination of urinary iodine alone gives inadequate information about the capacity of an individual to utilize an available iodine supply and it also shows the existence of different patterns of thyroid response during pregnancy. The history of iodine availability prior to and during pregnancy seems to be an important determinant of the mechanism of thyroid gland response to ensure the extra iodine needed by the growing fetus. Sponsorship: Financial support for this work was provided by the Swedish Medical Research Council and InDevelop AB, Uppsala University.