Objective: Selenium (Se) in the form of selenocysteine is an essential component of the family of the detoxifying enzymes glutathione peroxidase (Gpx) and of the iodothyronine selenodeiodinases that catalyse the extrathyroidal production of tri-iodothyronine (T-3). Thus, Se deficiency may seriously influence the generation of free radicals, the conversion of thyroxine (T-4) to T-3 and the autoimmune process. Therefore, we performed a randomised, placebo-controlled prospective study to investigate the effects of Se treatment on patients with autoimmune thyroiditis (AIT). Design and methods: Sixty five patients aged 22-61 years (median age 48 years) with AIT were recruited into two groups. Group I (Gr I) (n = 34) was treated with selenomethionine (Seme) 200 mug, Plus L-thyroxine (LT4) to maintain TSH levels between 0.3-2.0 mU/l, whereas group II (Gr II) (n = 3 1) received LT4 plus placebo over a period of 6 months. Moreover, the pharmacokinetics of Seme were studied in 10 patients and eight volunteers at baseline and 2 h, 4 h, 6 h and 24 h after oral administration of a 200 Kg tablet of Seme. Finally, Se levels were measured at the end of the study in some patients of both groups and their results were correlated with thyroid hormone levels. Results: In the pharmacokinetics study, basal serum concentration of Se (75+/-6 mug/l) was within the reference range (70-125 mug/l), it promptly increased at 2h, peaked at 4h (147+/-17 mug/l; P < 0.0001) and it was abundant in serum at 24h. In Gr 1, antibodies against thyroid peroxidase (anti-TPO) levels showed an overall decrease of 46% at 3 months (from 1875+/-1039U/l to 1013+/-382U/l; P < 0.0001) and of 55.5% at 6 months. In Gr 11 the overall decrease of anti-TPO amounted to 21% at 3 months and to 27% at 6 months (from 1758+/-917 U/l to 1284+/-410 U/l; P < 0.005). There were no significant changes of antibodies against thyroglobulin levels between the groups. At the end of this study Se levels were found to be statistically significantly increased in Gr I (n = 9/34) compared with Gr II (n = 11/31) (97+/-8.4 vs 79+/-8; P < 0.01) but no correlation with thyroid hormone was found. Conclusions: Seme is proven to be rapidly absorbed by the gastrointestinal tract. It appears to be useful as adjunctive therapy with LT4 in the treatment of AIT. The exact mechanism(s) is not very well determined, it might enhance the activity of detoxifying enzymes and enforce the defense against oxidative stress.