The hygiene hypothesis suggests that improved hygiene, and an associated reduction in exposure to infections in early life, increases the risk of development of allergic disorders in later childhood. This thesis reports the first phase of a longitudinal study designed to examining this hypothesis, using data from the Millennium Cohort Study (MCS). Objective measures of exposure to infections were obtained using oral fluid samples collected from 14 630 children in the home setting when cohort children were age 3 years. Samples were tested for antibodies to Epstein-Barr Virus (EBV), Varicella-Zoster Virus (VZV) and total IgG as a marker of sample quality. Oral fluid samples were received from 11883 (81.3%) cohort children, over 98% of which were suitable for assay testing. Children whose mothers were of Black Caribbean ethnicity or who lived in non-English speaking households were less likely to provide a sample. Mothers reporting a history of asthma were more likely to return a sample from their child. The estimated prevalence of EBV in the UK was 21.5% (95% Confidence Interval (CI): 20.4-22.5). Risk factors for early acquisition were maternal Black African, Pakistani, Bangladeshi, Other or Indian ethnic group, with decreased acquisition amongst children with mothers aged over 40 years. The prevalence of VZV in children in the UK was estimated using oral fluid assay and maternal report of varicella history, with prevalence estimates of 47.3% (95% CI: 46.1-48.6) and 46.1% (95% CI: 44.8-47.4) respectively. The level of concordance between measures was 70% with Positive Predictive Value (PPV) 68.0% and Negative Predictive Value (NPV) 70.6%. Oral fluid sample quality was evaluated by measurement of total IgG concentration with median 2.19 mg/L (IQ range: 0.99-4.4). High total IgG was associated with children whose mothers were of Bangladeshi ethnicity or with shorter duration of postal transit time. Oral fluid is a highly acceptable and feasible biological sample for collection in a large-scale child cohort study, although formal interpreter support may increase participation. Ethnic, social and demographic patterns of acquisition may be mediated by cultural, behavioural or biological factors. These findings form the basis for informing future studies of oral fluid collection, and together with further work regarding oral fluid sample quality will give the most recent prevalence estimates for EBV and VZV in the UK.