Studies were performed to explore the mechanism underlying the impaired generation of 125-I-3,5,3'-triiodothyronine (T3) from 125I-thyroxine (T4) (T3-neogenesis)) in preparations of liver from rats fasted for 48 h and the prevention of this effect by the feeding of glucose. T3-neogenesis in livers from fasted animals and those fed chow or glucose was assessed in various mixtures of crude microsomal fractions with either buffer or cytosols. T3-neogenesis was mediated by an enzyme present in the microsomal fraction whose activity was enhanced by cytosolic cofactor(s). In livers from animals fasted for 48 h, the supporting activity of cytosol was decreased, whereas the activity of the enzyme was unaffected. Administration of glucose as the sole nutritional source prevented the decrease in the supporting activity of hepatic cytosol that was regularly observed in the case of animals totally deprived of food. The diminished supporting activity for T3-neogenesis provided by liver cytosol from fasted animals was restored to normal by enrichment with either NADPH or GSH, but the two cofactors appeared to act at different loci. GSH stimulated T3-neogenesis in microsomes incubated in the absence of cytosol, i.e., in buffer, whereas NADPH did not. The stimulatory effect of both agents was blocked by the sulfhydryl oxidant, diamide, which also inhibited T3-neogenesis in mixtures of microsomes with cytosols. Taken together, these observations suggest that GSH acts directly on the enzyme in the crude microsomal fraction, whereas NADPH acts within the cytosol, possibly by increasing the concentration of GSH through the action of the enzyme glutathione reductase, for which NADPH is a cofactor. In this light, the decreased supporting activity of hepatic cytosol from starved animals appears to reflect, at least partly, a decreased concentration of one or both cofactors. The direct stimulation of enzyme activity by GSH, and the apparent lack of inhibition of unstimulated activity by diamide, suggests that the 5'-monodeiodinase for thyroxine that mediates T3-neogenesis may be a GSH transhydrogenase.