In his famous 1962 book, The Structure of Scientific Revolutions, Thomas Kuhn observed that scientific fields undergo periodic paradigm shifts, which open up new approaches to understanding that would never have been considered valid before. For example, currently it seems very clear to state that thyroid hormone (T3) exerts its activity directly on gene expression by binding to nuclear receptors, which are ligand-dependent transcription factors. However, during the 1970s, this view was iconoclastic because hormones were believed to signal only at the cell surface by binding to membrane receptors. However, the cloning of a first gene, now called THRA, which encodes the thyroid hormone receptor (TR)1 nuclear receptor (1) followed by the in vitro characterization of the receptor, definitely established the existence of the nuclear pathway for T3 action, also called the genomic pathway. This was followed by the cloning of the THRB gene, a paralog of THRA, which encodes two other nuclear receptors, TR 1 and TR 2. These two receptors, collectively called TR , differ at their N terminus but both display high similarities with TR 1. Interestingly, due to variations in promoter usage, splicing, and translation initiation THRA was later found to encode a number of other proteins, which do not act as nuclear receptors and whose functions still remain uncertain. Thus, the cloning of THRA and THRB established a new paradigm that demonstrated that some hormones like T3 directly bind transcription factors in the cell nucleusand regulate target gene transcription in contrast to peptide hormones that bind to cell membrane receptors. Nevertheless, a nonnuclear action of T3 has never been ruled out. Some feel that the broad spectrum of physiological functions of T3 would better fit with the existence of more than one mode of action. Importantly, some responses to T3, such as the regulation of movement of ions, sugars, and amino acids, are probably too rapid to be compatible with the time needed for RNA synthesis. Over the years, many different studies have illustrated the possibility for a rapid response and several alternative pathways, called nongenomic pathways for T3 action have been proposed (reviewed in reference 2). This term, however, is confusing because some of the proposed pathways ultimately result in the transcriptional activation of nuclear (3) or mitochondrial (4) genes. Furthermore, the accumulation of in vitro data based on biochemistry and cell culture is not sufficient to demonstrate the developmental or a physiological relevance of the nongenomic pathways, and the matter remains highly controversial. The main weakness of most of these studies is that they lack genetic support. Most importantly, THRA/THRB knockout mice devoid of T3 nuclear receptors have been produced 15 years ago (5, 6), and no residual response to T3 stimulation has been reported in these mice up to now. Therefore, if alternate T3 signaling pathways exist, they probably require the presence of some of the THRA or THRB gene products. If this assumption is correct, the distinction between genomic and nongenomic pathways by genetic means becomes problematic. Among the proposed nongenomic pathways, one relies on a direct interaction between TR and phosphatidylinositol 3-kinase (PI3K). This nongenomic action is suggested to explain the presence of follicular thyroid carcinoma in mice with a frame-shift mutation in the TR C-terminal helix of TR . This so-called TR PV mutant