The Moxon and Murphy (1978) model proposes that bacterial infection is initiated by invasion of a small group of closely related organisms at a single point.  The organisms expand to produce a homogeneous population and over time diversity is generated.  Subsequent selective pressures reduce the population to organisms with characteristics that assist survival in that particular niche.  Diversity is produced by numerous mechanisms, and in an infection the ability to generate variation may be the most crucial characteristic of all. To test the original infection model, which was proposed from animal work, analysis of the diversity of bacterial isolates from three different patients was carried out.  In this PhD the diversity of antibiotic resistance mechanisms in Pseudomonas aeruginosa serial isolates from a single patient was investigated.  Also, the diversity of Escherichia coli populations from another patient was explored.  Finally, the mutation frequency in these E. coli populations and also in a set of serial antibiotic resistant E. coli isolates (from a previous study;  Low et al., 2001) was examined. The data from the different isolate sets provided some evidence to support the Moxon and Murphy infection model.  The P. aeruginosa data showed diversity in antibiotic susceptibility and b-lactamases.  The results also indicated that the strains had not evolved sequentially suggesting the occurrence of separate infection pockets within the model.  The E. coli population isolates showed strain and sub-strain variation, giving strong confirmation that diversification occurs within the body.  Finally, the antibiotic resistant E. coli isolates showed increasing mutation frequency, and may provide an example of the generation of a mutator phenotype within the host under strong selective pressure.