The identification of four sortase homologues srtA-D in the pneumococcal genome prompted their investigation as candidate pneumococcal virulence factors. Not all of these sortase genes were present in both sequenced strains and so sortase gene distribution was investigated among a collection of clinical isolates. In contrast to srtB, C and D, srtA was found in all strains examined and so was selected for further study. It was subsequently found to contribute to pneumococcal virulence in mouse models of pneumonia, bacteraemia and colonisation. This is the first demonstration of a contribution of srtA to pneumococcal virulence. To complement this examination of srtA, some of the surface proteins known or likely to be anchored by SrtA were also investigated for a role in virulence in the animal model of pneumonia. In addition, two genes, annotated in the pneumococcal genomes as a macrophage infectivity potentiator protein and an exfoliative toxin A were also investigated and found to be novel pneumococcal virulence factors. However, it appears they have been incorrectly annotated and these genes do not represent a macrophage infectivity potentiator protein and exfoliative toxin A. Instead, one of them seems to be involved in the response to oxidative stress while no function for the other can yet be ascertained. Janus mutagenesis is a novel technique for manipulation of the pneumococcal genome allowing the creation of mutations that lack a selectable marker. This provides an accessible and potentially powerful method to easily alter the genome to make informative mutations. This thesis describes the first use, to our knowledge, of Janus mutagenesis to investigate pneumococcal virulence.