Shigellosis (bacillary dysentery) is a leading cause of diarrhoeal deaths globally, caused by lineages of Escherichia coli that convergently evolved to become specialised human pathogens. The independent acquisitions of a large virulence plasmid (pINV) by Shigella spp. and enteroinvasive E. coli (EIEC) conferred the ability to invade epithelial cells and manifest severe diarrhoeal disease. Within Shigella and EIEC, some lineages are significantly more epidemiologically successful than others, but factors underlying their success remain underexplored. This thesis combines genomics, bioinformatic approaches and the use of both in vivo and in cellulo infection models to explore the pathogenicity of different subgroups of Shigella to identify factors that contribute to their epidemiological success in humans. Firstly, I use the recently emerged serotype O96:H19 (and sequence type (ST) 99) EIEC as a model to explore the early stages of evolution. By reconstructing a dated phylogeny of ST99 E. coli, distinct pINV-positive and -negative clusters were identified and subsequent zebrafish infection revealed distinct mechanisms of virulence within these two clusters. This work provides novel insights into the keystone role of pINV acquisition in the virulence of a recently emerged clone. Next I focus on Shigella sonnei, a more established but relatively recently diverged Shigella subgroup emerging as the dominant agent of shigellosis worldwide. A collection of epidemiologically relevant clinical isolates was curated and underwent whole genome sequencing to generate fully completed genomes; comparative genomics was then performed to characterise S. sonnei lineages and identify lineage-dependent genomic variation. This analysis revealed ongoing adaptive evolution within S. sonnei, characterised by the accumulation of insertion sequences, pseudogenisation and structural rearrangements. Finally, I used a wet-lab approach to characterise different lineages of S. sonnei in vivo (using zebrafish) and in cellulo (using HeLa cells and human neutrophils), and then experimentally explore factors that contribute to their virulence. Here, the overall aim is to uncover functional variations contributing to differences in S. sonnei pathogenicity and epidemiological success. Overall, this interdisciplinary approach links the epidemiological landscape of Shigella to both genome content and pathogenicity in vivo for the first time, providing an innovative pipeline to study the evolution of bacterial pathogens and uncover signatures of pathogen success.