The sporulation-related gamma-D-glutamyl-(L)meso-diaminopimelic-acid-hydrolysing peptidase I of Bacillus sphaericus NCTC 9602 has been analysed by proton-induced X-ray emission. It contains 1 equivalent Zn2+ per mol of protein. As derived from gene cloning and sequencing, the B. sphaericus Zn peptidase I is a two-module protein. A 100-amino-acid-residue N-terminal domain consisting of two tandem segments of similar sequences, is fused to a 296-amino-acid-residue C-terminal catalytic domain. The catalytic domain belongs to the Zn carboxypeptidase A family, the closest match being observed with the Streptomyces griseus carboxypeptidase [Narahashi (1990) J. Biochem. 107, 879-886] and with the family prototype, bovine carboxypeptidase A. The catalytic domain of the B. sphaericus peptidase I possesses, distributed along the amino-acid sequence, peptide segments, a triad His162-Glu165-His307 and a dyad Tyr347-Glu366 that are equivalent to secondary structures, the zinc-binding triad His69-Glu72-His196 and the catalytic dyad Tyr248-Glu270 of bovine carboxypeptidase A respectively. The N-terminal repeats of the B. sphaericus peptidase I have similarity with the C-terminal repeats of the Enterococcus hirae muramidase 2, the Streptococcus (now Enterococcus) faecalis autolysin and the Bacillus phi PZA and phi 29 lysozymes, to which a role in the recognition of a particular moiety of the bacterial cell envelope has been tentatively assigned. Detergents enhance considerably the specific activity of the B. sphaericus peptidase I.