Objective: To longitudinally evaluate the timing of maternal thyroid underfunction occurrence in mildly iodine-deficient (ID) pregnant women, and ultimately assess the benefit of thyroid function testing at early gestation only in identifying maternal thyroid underfunction. Participants/methods: Serum free-thyroxine and TSH were measured in 220 consecutive women once in early pregnancy (by week 12) and twice per trimester subsequently. Anti-thyroperoxidase and antithyroglobulin were also determined at initial and final observation. Results: Thyroid autoantibodies were detectable in 8.2% women. Overall, the prevalence of hypothyroidism over the course of gestation was 11.8% (26/220), with a relative risk of hypothyroidism in antibody-positive women of 5.0 (c 2 20.02, P!0.0005). Nonetheless, almost 70% hypothyroid women tested negative for thyroid autoantibodies. Fifteen/26 (57.7%) hypothyroid women were identified at presentation, and the remaining 11 at either early (6/11) or late (5/11) phases of the 2nd trimester. Isolated hypothyroxinemia was observed in 56/220 (25.4%) women, mostly from the 2nd trimester onwards. Conclusions: In mildly ID areas thyroid function testing early in gestation seems to be only partly effective in identifying thyroid underfunction in pregnant women. Indeed, in our series more than 40% hypothyroid women would not have been diagnosed had we limited our observation to early thyroid function tests alone. Although thyroid autoimmunity carried a 5-fold increased risk of hypothyroidism, iodine deficiency seems to be a major determinant in the occurrence of thyroid underfunction. Adequate iodine supplementation should be strongly recommended to meet the increased hormone demand over gestation.