Beta-Lipotropin was purified from human pituitary glands to a purity of greater than 90%. The amino acid compositions of beta-lipotropin and its three cyanogen bromide cleavage peptide fragments were in agreement with the structure proposed by Li and Chung [Li, C.H. & Chung, D. (1981) Int. J. Pept. Protein Res. 17, 131-142]. However, the amino acid sequence of its NH2-terminal 46 amino acid residues established here differs both from the sequence derived from the direct sequence analysis of the peptide reported by Li and Chung and from that predicted on the basis of the nucleotide sequence of the human pro-opiolipomelanocortin gene proposed by Chang et al. [Chang, A.C.Y., Cochet, M. & Cohen, S.W. (1980) Proc. Natl. Acad. Sci. USA 77,4890-4894] but agrees with the structure recently derived by direct sequence analysis by Hsi et al. [Hsi, K.L., Seidah, N.G., Lu, C.L. & Chrétien, M. (1981) Biochem. Biophys. Res. Commun. 103, 1329-1335] and predicted on the basis of nucleotide sequence analysis by Takahashi et al. [Takahashi, H., Teranishi, Y., Nakanishi, S. & Numa, S. (1981) FEBS Lett. 135, 97-102]. These discrepancies, found from residues 9 to 25 of beta-lipotropin, could result from pro-opiolipomelanocortin gene polymorphism, from the existence of multiple genes for pro-opiolipomelanocortin, or, more probably, from minor errors in nucleotide and amino acid sequence analyses.