In the brain, 5′‐deiodinase (5′‐D) is responsible for the metabolic activation of thyroxine (T₄) into 3,5,3′‐triiodothyronine (T₃) and 5‐deiodinase (5‐D) deiodinates T₄ and T₃ into inactive metabolites. This study examines the effects of factors known to induce astroglial 5′‐D activity on the 5‐D activity in cultured rat astroglial cells. The potencies of these factors were compared after 8 h of incubation, when stimulations by these factors near their maximal effects. 12‐O‐Tetradecanoylphorbol 13‐acetate (TPA) at 10⁻⁷M was a potent inducer of 5‐D activity, producing a 30‐ to 80‐fold increase after 8 h. The maximal effect of TPA was observed after about 14 h. The TPA stimulation of 5‐D activity was not dependent on glucocorticoids, unlike 5′‐D activity. In comparison with TPA, 8‐bromo‐cyclic AMP (10⁻³M) was a poor inducer of 5‐D activity whereas it is an excellent inducer of 5′‐D activity. It produced a 2‐ to 20‐fold increase in 5‐D activity after 8 h. Natural acidic fibroblast growth factor (20 ng/ml) produced a degree of stimulation similar to that of TPA after 8 h. The maximal effect of acidic fibroblast growth factor was observed after about 16 h (until a 120‐fold increase). Recombinant acidic fibroblast growth factor also induced 5‐D activity. Basic fibroblast growth factor was less potent than acidic fibroblast growth factor for increasing 5‐D activity (maximal increase by 40‐ to 50‐fold after 8 h). Platelet‐derived growth factor (20 ng/ml) and epidermal growth factor (100 ng/ml) were poor inducers of 5‐D activity (8‐ to 12‐fold increase at 8 h); insulin (10⁻⁶M) was without effect. The 5‐D activity induced by TPA and acidic fibroblast growth factor manifested the characteristics of type III 5‐D (K_m for T₃ 0.5–0.8 nM, thiol‐dependent, 6‐n‐propyl‐2‐thiouracil‐insensitive). The results demonstrate that, like 5′‐D activity, 5‐D activity is induced by multiple pathways. The relative potencies of TPA, fibroblast growth factors, and cyclic AMP on 5‐D and 5′‐D activities were different, as were the time courses of their actions. These data indicate that 5‐ and 5′‐D are distinct enzymes and support the view that T₃ availability may be controlled not only by regulating T₃ production, but also by regulating T₃ and T₄ degradation.