The effects of Se on thyroid metabolism in a New Zealand population are investigated, including (a) the relationship between Se and thyroid status, and (b) the effect of Se supplementation on thyroid status. The data used come from two cross-sectional studies of Se, I, thyroid hormones and thyroid volume (studies 1 and 4), and three Se intervention studies in which thyroid hormones, Se and glutathione peroxidase (GPx) activities were measured (studies 2, 3 and 5). There were no significant correlations between Se status and measures of thyroid status after controlling for sex at baseline or after supplementation in any of the studies. When data from study 4 were divided into two groups according to plasma Se, plasma thyroxine (T4) was lower in males with higher plasma Se levels (P=0·009). Se supplementation increased plasma Se and GPx activity, but produced only small changes in plasma T4 and triiodothyronine (T3):T4 ratio. In study 2, there was a significant reduction in plasma T4 (P=0·0045). In studies 3 and 5 there were small decreases in plasma T4 and a small increase in the T3:T4 ratio, which were not significantly different from placebo groups. Lack of significant associations between plasma Se and thyroid status, and only small changes in T4 suggest that Se status in New Zealand is close to adequate for the optimal function of deiodinases. Adequate plasma Se may be approximately 0·82–0·90 μmol/l, compared with 1·00–1·14 μmol/l for maximal GPx activities.