Pakistani women in the UK have high levels of mental illness, alongside low levels of outpatient mental health service use, compared with women of other ethnic groups. Further, previous studies have suggested that Pakistani women have particularly low levels of social support, and high levels of social isolation which may reduce their chances of coming into contact with mental health services. However, to date, there has been little empirical evidence to support this. This thesis investigated the mental health service use, social networks' structure and function, and the relationship between the two, for Pakistani women compared with women of other ethnic groups. This was done using a systematic review of the relevant literature, and statistical modelling using two large nationally representative datasets from the UK. The first dataset, Understanding Society, was used to formulate latent classes of support networks, subsequently used in regression models to compare the support available in Pakistani women's networks with women of other ethnic groups. The second dataset, Ethnic Minority Psychiatric Illness Rates in the Community (EMPIRIC), was used to ascertain the influence of social networks (perceived social support, contact with relatives and friends, network composition, and size) on the use of outpatient mental health services, using logistic regression modelling. These data were also used to build a structural equation model to test the direct and indirect effects of social networks on outpatient mental health service usage, via their impact on mental illness. Pakistani women (along with Bangladeshi women) had the lowest rate of mental health service use, compared with women in other ethnic groups. Further Pakistani women were more likely to be socially isolated than White majority women, but there were largely no differences between Pakistani women and other ethnic minority women in the structure and function of social networks. Finally, there was evidence to suggest that social networks indirectly reduced mental health service use via their impact on mental illness. There were only small ethnic differences in the indirect effect of social networks on mental health service use, and these differences did not explain Pakistani women's under-use of mental health services.