The prevalence of rheumatic heart disease (RHD) in the Aboriginal population of the Australian Northern Territory is high, and Streptococcus pyogenes skin infections likely contribute to this. A promising candidate S. pyogenes “30mer” vaccine is composed of 30 pharyngitis associated type-specific antigens from the S. pyogenes M protein. Cross opsonisation experiments suggest that 30mer vaccine protection may extend to non-cognate emm types. A new “emm cluster” scheme for classifying M protein is based on the full-length coding sequence, and correlates with functional and immunological properties, and anatomical tropism. Twenty-seven years of research in the Northern Territory has yielded 1810 S. pyogenes isolates with clinical and emm type data. The primary aim was to analyse these data with reference to the emm cluster scheme and cross opsonisation information, to inform estimation of 30mer vaccine efficacy in the Northern Territory. The isolates encompass 101 emm types. Variants of cluster A-C were enriched in throat isolates, and variants of emm cluster D enriched in skin isolates. Throat isolates were enriched for 30mer vaccine cognate emm types in comparison with skin isolates of which only 25% were vaccine emm types. While cross opsonisation data indicates potential for enhancing 30mer vaccine coverage, more than one third of skin isolates were within 38 emm types untested for cross opsonisation. Emm cluster D variants, in particular emm cluster D4, were not only all non-cognate with the vaccine, but were abundant and diverse, and less likely to be cross-opsonisation positive than other emm clusters. Long term persistence of many emm types in the study area was revealed. It was concluded that the 30mer vaccine efficacy in the Northern Territory will likely require both cross protection, and additional measures to elicit immunity against variants of emm cluster D.