Female preference expressed through mate choice for attractive males represents a fundamental mechanism for the selection and propagation of heritable advantageous traits in animal populations. The presence of pathogenic infection in males can alter female mate choice, commonly rendering affected males less attractive than uninfected conspecifics, although there are several exceptions. It has been suggested that in wild populations, variation between males in their ability to cope with infection may be reflected in differences in signalling. Signals may include secondary sexual characteristics, behaviour, or other changes associated with clinical pathology. Recognition of these signals could provide a reliable estimate of heritable resistance to locally relevant pathogens, as highly resistant males may invest more heavily in signals of attractiveness or suppress indicators of pathology, and so experience greater reproductive success. Heritable resistance in the context of local pathogens is likely to increase offspring survival, so improving the reproductive success of both the male and female parent. Interpretation of female preference behaviour from an ecological perspective demands that experimental design takes into account the evolutionary history of the host-pathogen combination under study. Natural host-pathogen combinations are likely to be mutually adapted through co-evolution, resulting in optimisation of avoidance and transmission strategies respectively. Thus observations gained through experimental use of naturally occurring pathogens in wild study populations are more likely to elicit ecologically relevant responses than non-natural pathogens. A serological survey during this project found variable prevalence of several viral pathogens in wild and wild-derived captive populations of house mice (Mus domesticus). This survey also found that most viruses showed low or no transmission in captivity, with the exception oflymphocytic choriomeningitis virus (LCMV). As this pathogen is an important zoonosis, the captive population was screened to eliminate LCMV and thus remove any associated risk to personnel. Transmission of LCMV occurred vertically (to the foetuses in utero) in most cases, and several animals that were persistently infected with virus had no detectable antibody to this pathogen. This has important implications for standard laboratory screening for LCMV, where the use of serological methods alone could fail to detect persistent infections. The effects of immune stimulation and viral infection on male competitive behaviour and female preference for males and their scents were investigated. Previous studies had demonstrated infectionassociated changes to urinary scent signalling. Thus urine of control, vaccinated and infected animals was subjected to biochemical analysis. Immune stimulation was achieved by vaccinating animals with the novel antigen keyhole limpet haemocyanin. During infection studies, animals were inoculated with mouse adenovirus, a natural pathogen of wild mice that had been identified during the serology survey. Immune stimulation caused small a reduction in male scent marking behaviour, but did not alter female preference or urine biochemistry. Infection caused no change to male urine marking behaviour, but reduced urinary protein concentration. This did not alter female preference. Pregnancy block or the 'Bruce' effect may represent a mechanism females could use to avoid the deletenous effects of infection on offspring. Laboratory studies have demonstrated that female mice will termmate gestation if exposed to an unfamiliar male within a limited time after mating. As these studies ofte.n use inbred laboratory strains, males are classed as 'unfamiliar' if they arw of a different inbred stram to the stud male. However, scent changes associated with infection may provide a different source of unfamlhanty. The 'Bruce' effect is traditionally assumed to have evolved as it offers potential advantages.to both males and females. However, we show that female mice are able to control whether pregnancy is blocked or maintained by altering their own exposure to male scent at times critical for the activation of this physiological mechanism. Thus females may control'the 'Bruce' effect without reference to male advantage, and this behaviour may have evolved solely to improve female reproductIve success.