Following the emergence of a new variant of Creutzfeldt-Jakob disease (vCJD) 6 years ago, and the gradual rise in clinical cases, there has been increased speculation regarding the overall magnitude of this epidemic in Great Britain. In this paper, we explore the epidemiological factors and uncertainties determining the scale of this epidemic in light of the most recent data on reported vCJD mortality. Our results demonstrate that, while the magnitude of the uncertainty has decreased dramatically since 1996, it is still not possible to predict with any degree of accuracy the final magnitude of this epidemic, with the 95% confidence interval for future cases being from 10 to 7000 deaths. However, short-term projections show that it is unlikely that a dramatic increase in case numbers will be observed in the next 2-5 years (95% confidence interval for 2 years: 10-80 cases, for 5 years: 10-200 cases). The results confirm significant age-dependent susceptibility/exposure to infection, with the likelihood profile demonstrating that those aged between 10 and 20 years are at highest risk of infection. We also demonstrate how projections based on onset data may be substantially biased, and explore the sensitivity of results to assumptions concerning the exposure to bovine spongiform encephalopathy (BSE) and the incubation-period distribution.