Enterohaemorrhagic Escherichia coli (EHEC) strains are associated with gastrointestinal and severe systemic disease in humans. EHEC O157:H7 is the most common serotype causing human infections in North America and the UK. Human infections mainly originate from cattle, through either direct contact with infected animals or indirectly through contamination of food or water with animal faeces. From the sequencing of EHEC O157 strains, it is clear that the genomes contain multiple prophages, many of them cryptic, which define this E. coli pathotype. These regions include the locus of enterocyte effacement (LEE) which is a critical horizontally acquired pathogenicity island and encodes a type III secretion system (T3SS). The T3SS translocates effector proteins into epithelial cells that enable tight attachment to these host cells and also modify innate responses and other cellular functions to promote persistence in the animal host. The T3SS is essential for the colonisation of cattle by EHEC O157 where it is localised to the terminal rectum. The regulation of T3S is complex with many regulators and environmental factors already identified. Previous work has demonstrated marked variation in the levels of T3S among EHEC O157 strains. The aim of this research was to further investigate the regulation of T3S towards two objectives: (1) to understand the localisation of EHEC O157 at the terminal rectum of cattle; (2) to understand the strain variation in T3S. (1) In relation to rectal and mucosal colonisation, established aerobic/anaerobic regulators were investigated including arcA, fnr, narX, narQ. Briefly, arcA, fnr, narX, narQ were deleted in an E. coli O157 strain ZAP198 by lambda red recombination. Apart from the fnr mutant which showed lower levels of T3S, the remaining mutants displayed similar T3S protein levels compared to the wild type strain. In addition, no significant changes in adherence and A/E lesion formation capacity were measured for the mutants following interaction with bovine epithelial cells. (2) Strain secretion variation was approached in two ways; the first was to control expression from the LEE1 operon, required for T3S expression, in order to both induce expression and examine the importance of downstream regulation. The second was to investigate variation in T3S between different phages types of EHEC O157. While attempts to construct an inducible T3SS were not successful, intermediate strains made in the process have been useful to dissect how regulators being studied in the laboratory control T3S. The main novel insights from the research have come from examining T3S in different EHEC O157 phage types. We found that the average level of T3S in PT 21/28 strains was lower than in PT 32 strains. Interestingly, most (90%) of PT 21/28 strains contained both Stx2 and Stx2c phages. In contrast, only 28% of PT 32 strains had both phages. Taken together, this raised the possibility that Stx phage integration might have a repressive impact on T3SS regulation in E.coli O157:H7. This hypothesis was addressed using a number of different approaches. Deletions of Stx phages were constructed and these had increased levels of T3S when compared to the parental strains. This phage regulation of T3SS was confirmed in an E. coli K12 background by examining an induced LEE1 reporter in the presence and absence of a transduced Stx2 phage. In addition, it was shown that deletion of the CII phage regulator led to increased T3S and may contribute to the Stx phage repression reported above. This work demonstrates for the first time that Stx phage integration represses T3S expression. It is proposed that this control may limit immune exposure of this critical colonisation factor and that the repression actually allows activation by prophage encoded regulators, including PchA/B, that co-ordinate T3S and non LEE-encoded effector expression to promote epithelial cell colonisation.