The development of a group A streptococcal (GAS) vaccine has been hindered by a number of hurdles including the high serotypic diversity of the pathogen, autoimmune complications following repeated GAS infections, and the lack of non-murine models to validate proposed GAS vaccine formulations. To overcome these hurdles, we generated a global GAS genome database to unravel the evolutionary dynamics of this major human pathogen. This database of 2,083 genomes were obtained from 22 countries, with a focus on sampling from streptococcal endemic settings. We identified a core panel of pre-clinical GAS vaccine antigens that would provide theoretical global coverage on the basis of >99% antigen carriage and <2% sequence heterogeneity. Using this platform, 5 conserved antigens (arginine deiminase [ADI], C5a peptidase [SCPA], streptolysin O [SLO], interleukin-8 [IL-8] protease [SpyCEP], and trigger factor [TF]), that have not been linked to autoimmune complications yet are highly conserved within a global context, were investigated as a putative multi-component vaccine formulation. We developed a non-human primate (NHP) infection model of GAS pharyngitis and evaluated the protective efficacy of the 5 conserved antigen formulation termed ‘Combo5’. Antibody responses against all Combo5 antigens were detected in NHP serum, and immunised NHPs showed a reduced pharyngitis and tonsillitis compared to controls. Within an evolving global bacterial pathogen such as GAS, we have identified a number of proposed pre-clinical GAS vaccine antigens that fulfil the criteria for a global vaccine and provide protection from pharyngitis in a NHP model. Our work establishes a technical and experimental framework for the development of a global GAS vaccine.