Prion diseases are rare, invariably fatal, neurodegenerative diseases, occurring in sporadic, genetic and iatrogenic forms in animals and humans, for which there is no aeceptahlc diagnostic test in life and no effective treatment. In humans the commonest prion disease is Creutzfeldt-Jakob Disease (CJD). Systematic prospective public health surveillance (PHS) of CJD was initiated in the UK in 1990 in response to the detection of a novel prion disease in cattle, bovine spongiform encephalopathy (BSE). The aim of PHS was to detect any change in the elinico-pathological phenotypc of CJD that could be attributable to human exposure to BSE. In this thesis I present a series of studies that evaluate various aspects of the PHS of CJD in the UK, 1990 - 2006. From 1990 to 2006, 2154 suspect CJD cases were referred to the National CJD Surveillance Unit (NCJDSU); 57% had a clinical or neuropathological diagnosis of CJD. Sporadic CJD (sCJD) accounted for the majority of cases. Age adjusted sCJD incidence increased over time in association with an increasing use of CSF 14-3-3 protein for case classification. Genetic prion disease accounted for 9.4% of all cases; 54 iatrogenic CJD cases mediated by recognised routes of transmission were identified. Variant CJD (vCJD), a novel human prion disease, was characterised by the NCJDSU in 1996; by 2006 there had been 165 incident cases in the UK. The primary vCJD epidemic peaked in the UK in 2000 (27 incident cases) and has been in decline since. Secondary transmission of vCJD through the transfusion of labile blood components has been identified, occurring during an asymptomatic phase of illness. The prevalence of asymptomatic vCJD infection in the population is not known. In characterising vCJD and contributing to establishing an aetiological link with BSE, the NCJDSU met a primary aim of PHS. In an evaluation the NCJDSU was found to be flexible, acceptable, sensitive, timely and representative. Falling post mortem rates and an increasing reliance on clinical diagnostic criteria, with evidence of sub-optimal and differential use of investigations to support a diagnosis of sCJD and vCJD are of concern, as is the rising positive predictive value of the system in the face of falling referral rates. NCJDSU operational criteria for the assessment of EEGs for case classification in sCJD were prospectively validated. The sensitivity of EEG was low and specificity high; EEG remains a valuable non-invasive investigation in sCJD if used in conjunction with other diagnostic tools. With the establishment of systematic prospective PHS the reliance on death certificates to ascertain suspect cases has diminished. The sensitivity and specificity of a death certificate diagnosis of prion disease in the UK are high. Death certificate data provide valid estimates of prion disease mortality in the UK. Uncertainty around the epidemiology and pathogenesis of vCJD and the emergence of novel atypical prion diseases in animals which pose an as yet unknown threat to human health, provide the imperative to continue systematic prospective PHS of prion disease in humans in the UK for the foreseeable future. The NCJDSU is well placed to achieve this.