One hundred and eighty euthyroid pregnant women were selected at the end of the first trimester of gestation on the basis of biochemical criteria of excessive thyroid stimulation, defined as supranormal serum thyroglobulin (TG > 20 micrograms/L) associated with a low normal free T4 index (< 1.23) and/or an increased T3/T4 ratio (> 25 x 10(-3)). Women were randomized in a double blind protocol into three groups and treated until term with a placebo, 100 micrograms potassium iodide (KI)/day, or 100 micrograms iodide plus 100 micrograms L-T4/day. Parameters of thyroid function, urinary iodine excretion, and thyroid volume were monitored sequentially. Neonatal thyroid parameters, including thyroid volume by echography, were also assessed in the newborns from mothers of the three groups. In women receiving a placebo, the indices of excessive thyroid stimulation worsened as gestation progressed, with low free T4 levels, markedly increased serum TG and T3/T4 ratio. Serum TSH doubled, on the average, and was supranormal in 20% of the cases at term. Urinary iodine excretion levels were low, around 30 micrograms/L at term. The thyroid volume increased, on the average, by 30%, and 16% of the women developed a goiter, confirming the goitrogenic stimulus associated with pregnancy. Moreover, the newborns of these mothers had significantly larger thyroid volumes at birth as well as elevated serum TG levels. In both groups of women receiving an active treatment, the alterations in thyroid function associated with pregnancy were markedly improved. The increase in serum TSH was almost suppressed, serum TG decreased significantly, and changes in thyroid volume were minimized (group receiving KI) or almost suppressed (group receiving KI combined with L-T4). Moreover, in the newborns of the mothers in the two groups receiving an active treatment, serum TG was significantly lower, and thyroid volume at birth was normal. The effects of therapy were clearly more rapid and more marked in the group receiving a combination of T4 and KI than in the women receiving KI alone. The differences could be partly attributed to the slightly higher amount of iodine received by women in the combined treatment. However, the main benefits of the combined treatment were almost certainly attributable to the hormonal effects of the addition of L-T4. Furthermore, the study demonstrated that the administration of T4 did not hamper the beneficial effect of iodine supplementation. In conclusion, the present work emphasizes the potential risk of goitrogenic stimulation in both mother and newborn in the presence of mild iodine deficiency.(ABSTRACT TRUNCATED AT 400 WORDS)