Four histones, a, b, c, d from procyclic Trypanosoma brucei brucei, which show similarities with the amino acid composition of the core histones H3, H2A, H2B and H4, were isolated and cleaved with Endoproteinase Glu-C. The fragments were separated by FPLC reversed phase chromatography and a subset of the fragments (a5, a9, b6, c8, d3, d9, d11) was subjected to sequence analysis. A 54–71% identity was found in the sequences of the fragment c8 and the C-terminal half of H2B and of three fragments of protein d covering the N-terminal half as well as the C-terminal region of H4. The amino acid sequence of the fragment a9 showed a 57 and 54% identity with H3 sequences of Saccharomyces cerevisiae and Xenopus laevis. Neither the a5 nor the b6 sequence could be aligned with histone sequences of other eukaryotes. The significant differences of 21–48% between the T. b. brucei, histone sequences and those of calf thymus histones, which are more pronounced than the differences of Tetrahymena pyriformis and the higher eukaryote, resulted partially from replacements of amino acids with different properties and indicate specific patterns of histone–histone and/or histone–DNA contact sites in the nucleosome of T. b. brucei. These differences, together with the lack of a functional histone H1, may be sufficient to explain the lack of a salt-dependent formation of the nucleosome filament into the 30 nm fibre, which reflects alternative methods of organizing and processing the genetic information in the nucleus of the protozoan parasite and which may be of chemotherapeutic significance.