Antibodies specific for the acetylated forms of histone H4 (H4) were produced in rabbits with a synthetic peptide corresponding to the 18 N-terminal residues of tetra-acetylated H4 (i.e. with acetyllysine at positions 5, 8, 12 and 16). Specificity was determined by inhibition assays using four additional peptides, each acetylated at only a single site. Using an antiserum (R6) specific for the acetylation site at Lys-5 we have estimated the proportion of Lys-5 sites acetylated in the mono-, di- and tri-acetylated forms of H4 from randomly growing human HL-60 cells. The values obtained (7%, 29% and 61% respectively) differ from those expected if acetylation were random (i.e. 25%, 50% and 75%) or if site usage followed a set order for all H4 molecules (i.e. a jump from 0% to 100%). Antibodies from a second animal (R5) bound preferentially to peptides acetylated at Lys-12 and also bound to mono-acetylated H4 relatively weakly in several cell types. In contrast, mono-acetylated H4 from metaphase HeLa cells labelled more strongly with both antisera, indicating significant acetylation at Lys-5 and Lys-12. We conclude that (1) the sites at Lys-5 and Lys-12 are under-used in mono-acetylated H4 from a variety of mammalian cell types and Lys-8 and/or Lys-16 are therefore the first to be acetylated, (2) more than one order of site usage is possible and (3) there is a metaphase-specific shift in site usage. These results suggest that H4 acetylation plays a role in the modulation of chromatin structure in mammalian cells.