The H2A histone of the protozoan Tetrahymena pyriformis was isolated by Bio-Gel P-10 chromatography from the H2A+H3 fraction obtained on a large scale, as described previously [Nomoto, M. & Iwai, K. (1982) J. Biochem. 91, 719-723], and further purified by Sephadex G-100 chromatography. The purified H2A was shown to comprise approximately equimolar amounts of two variants, H2A (1) and H2A (2), differing in molecular weight. The H2A mixture was fragmented with cyanogen bromide, yielding one N-terminal fragment (101 residues), one middle fragment (17 residues), and two C-terminal fragments (19 and 14 residues). The N-terminal fragment, whose N-terminal was blocked, was sequenced by overlapping its tryptic peptides with the peptides derived from the fragment with three proteases and the tryptic peptides from citraconylated intact H2A. One of the citraconylated tryptic peptides showed the arrangement of the N-terminal, middle, and C-terminal fragments; the latter two fragments were directly sequenced by Edman degradation. Thus, the total sequences of H2A (1) and H2A (2) were completely determined; the two variants differ in the total residue number, 137 or 132, the molecular weight in the unmodified form, 14, 654 or 14, 126, His or Asn at residue 40, Ser or Thr at residue 124, and the C-terminal sequence at residues 128-137 or 128-132, respectively. These sequences are compared with the known H2A sequences, and the implications for the structure and function relationship of this histone species are discussed.