Antisera were raised to a tridecapeptide, Ser-Asp-Val-Thr-Lys-Arg-Gln-His-Pro-Gly-Arg-Arg-Phe, that was synthesized based on the sequence (residues 166-178) of a proposed cDNA for pro-TRH reported by Lechan et al. With this antiserum, immunostaining of Western blots of rat brain extracts revealed two major proteins with mol wt (Mr = 39,000 and 52,000) considerably larger than that of the largest protein (Mr = 29,000) that could be encoded by the cDNA of Lechan et al. Because these observations suggested the possibility of novel TRH precursors, we studied the immunocytochemical distribution of pro-TRH (39-52K) in rat brain. Our anatomical findings were 4-fold. 1) The distributions of 29K pro-TRH and 39-52K pro-TRH are not identical. 2) TRH is found only in regions containing 29K pro-TRH, 39-52K pro-TRH, or both. 3) There are regions that contain both 29K pro-TRH and 39-52K pro-TRH, but no TRH. 4) Regions containing only 39-52K pro-TRH do not contain 29K pro-TRH mRNA as mapped by Segerson et al. From these electrophoretic and anatomical observations, we postulate the existence of at least one and possibly two additional precursors that can be processed to TRH in rat brain.