Recent data have suggested that the iodothyronine, 3,5-diiodo-l-thyronine (T2), has selective thyromimetic activity. In vivo, T2 has been shown to suppress TSH levels at doses that do not produce significant peripheral manifestations of thyroid hormone activity. Furthermore, T2 has been shown to produce smaller increments in peripheral indices of thyroid status than does T3, when doses resulting in equivalent suppression of circulating TSH are compared. We have assessed the selective thyromimetic activity of T2 in vivo and in vitro, and performed in vitro studies to assess the potential molecular basis for these selective properties. T2 was 100-fold less potent than T3 in stimulating GH mRNA levels in GH3 cells. In contrast, the iodothyronines were almost equivalent in their ability to downregulate TRbeta2 mRNA levels in this cell line. Both 3,3'-diiodo-L-thyronine and thyronine exhibited no significant thyromimetic effects on either process. In vivo, doses of T2 and T3 that were equivalent in their induction of hepatic malic enzyme (ME) mRNA did not produce equivalent suppression of circulating TSH, with T2 being only 27% as effective as T3. T2 was up to 500-fold less potent than T3 in displacing [125I]-T3 from in vitro translated specific nuclear receptors (TRs) and GH3 cell nuclear extracts. Electrophoretic mobility shift assays, assessing the ability of T2 to produce dissociation of TRbeta1 homodimers from inverted palindrome T3 response elements, indicated that T2 was also 1000-fold less potent than T3 in this respect. These data confirm that T2 has significant thyromimetic activity, and that this activity is selective both in vivo and in vitro. However, there are no data to support a selective central effect, T2 being relatively more potent in stimulating hepatic ME mRNA than in suppression of TSH in vivo. The basis for this differential thyromimetic activity is not selective affinity of the different TR isoforms for T2, or divergent properties of T2 in competitive binding and functional assays in vitro.