Rheumatic fever and rheumatic heart disease remain very common in developing countries, and a vaccine to protect against these disorders would have a great impact on public health. A vaccine must target the M protein of group A streptococci (Streptococcus pyogenes), but until lately immunity was thought to be strain-specific and dependent on antibodies to the variable serotype-specific regions of the protein. Experiments in animals have suggested the conserved region of the M protein as a possible alternative target for protective antibodies. We constructed a 20-aminoacid peptide (peptide 145) within the conserved region of the carboxyl terminus of the protein. In mice the peptide induced serum antibodies that could opsonise reference type 5 streptococci. By enzyme-linked immunosorbent assay, positive responses to peptide 145 were obtained with serum from 77 (90%) of 86 Aboriginal subjects and 135 (81%) of 167 Thai subjects living in areas with high exposure to streptococci. Only 10 (14%) of 71 Caucasian subjects with low exposure to streptococci showed positive responses. There was no difference in the proportion positive between subjects with rheumatic heart disease and control groups (other or no heart disease). Antibodies to peptide 145 were able to opsonise isolates of streptococci from Aboriginal and Thai subjects with acute rheumatic fever as well as reference strains. This highly conserved part of the M protein may be a suitable target for vaccines to prevent streptococcal infections and their sequelae.