Objective To elucidate the persistence, or otherwise, of the pregnancy-related changes in the iodine metabolism and thyroid function in a population residing in an area of mild iodine de®ciency in the Sudan. Design: A longitudinal prospective cohort study involving pregnant women who were recruited during their third trimester of pregnancy and were followed up for up to nine months after delivery. Setting: The study was conducted among Sudanese women residing in the Omdurman area in Khartoum, an area with a total goitre rate of 17.5%. Subjects: Forty-seven pregnant women were recruited during their third trimester of pregnancy. Their mean age and weight were 29 AE 4.6 y and 62.4 AE 8.7 kg, respectively. Age matched healthy non-pregnant women living in the same area (n 40) served as a control group. Methods: Serum levels of thyrotropin (TSH), thyroxine (T4), triiodothyronine (T3), thyroglobulin (Tg) and the urinary iodine concentration (UIC) were determined during the third trimester (®rst occasion), and subsequently at three months (second occasion), six months (third occasion) and nine months after delivery (fourth occasion). Control subjects provided corresponding samples on one occasion. Main outcomes measures: UIC, TSH, Tg, FT4 and T3. Results: During the third trimester of pregnancy the median UIC and the free T4 (FT4) were lower than in the control group (P`0.0001, P`0.0001, respectively), while the median Tg was higher than in the controls (P`0.03). Three months post-partum maximum thyroidal stimulation was evident, with elevated serum levels of TSH and Tg (P`0.0001, P`0.03) and reduced UIC and serum T3 and FT4 (P`0.0004, P`0.0005, P`0.0001), compared with the control group. Nine months post-partum the thyroid function was restored to the pre-pregnancy state, and the median values of TSH, Tg, T3, FT4 and UIC did not differ signi®cantly from those in the control group. Conclusions: Our study suggests that the reversibility of the pregnancy-induced changes in the iodine status and thyroid function to the pre-pregnancy levels may depend on the iodine status of the mother during and after pregnancy. The thyroidal stress during the ®rst three months of the post-partum period, which is partially due to the iodine loss in the breast milk, justi®es further detailed studies to assess the iodine content of the breast milk and the role of breast milk as a vehicle for iodine supply of infants in situations of iodine de®ciency.