Understanding both the mechanistic basis of virulence and the evolutionary processes under which it can arise, is fundamental if we are to increase our knowledge of disease causing bacteria in an era of ever increasing antibiotic resistance. To date, there has been a substantial effort to understand virulence evolution both theoretically and experimentally. However, comparatively little experimental work has focussed on the evolution of virulence in opportunistic pathogens, and how virulence varies across multiple host organisms. In this thesis, the opportunist Pseudomonas aeruginosa was used to study virulence and its evolution in hosts. The virulence levels of different strains of P. aeruginosa, both laboratory and clinical, were tested in the nematode Caenorhabditis elegans. It was found that virulence of P. aeruginosa was lower in clinical strains isolated from chronic infections, and laboratory strains initially isolated from chronic infections, while a strain isolated from an acute infection was substantially more virulent. Using the C. elegans host, a selection experiment was carried out to test the effect of host density on virulence evolution. This resulted in a varied array of P. aeruginosa virulence phenotypes, but the worm evolved bacteria did not exhibit drastically different phenotypes from the worm negative control experiments. However, the degree of evolution observed across all treatment groups, compared with the wild type, highlighted the changes a bacterial population can undergo in vitro, even from a relatively short period of sub culture. Lastly, P. aeruginosa virulence was tested in alternative host organisms, Galleria mellonella (waxmoth larvae) and Vigna radiate (mung bean seedlings). Virulence varied according to the host, with virulence in G. mellonella being universally high, while no virulence was observed in V. radiate. This work highlights that high or low virulence cannot automatically be assumed in different hosts for opportunistic bacteria such as P. aeruginosa. Furthermore, the results highlight the unpredictable nature of virulence evolution, and how dependent the emergence of virulence is on the interaction between the pathogen and the host. A more complete understanding of the principles underlying bacterial evolution and virulence could allow for the construction of more specific and accurate studies, with experimental conditions carefully engineered to best replicate the clinical conditions of interest.