The prevention of congenital rubella syndrome (CRS), as a complication of rubella infection during pregnancy, is the main aim of rubella vaccination programmes. However, as vaccination of infants leads to an increase in the average age at which those who were not immunized become infected, certain rubella vaccination programmes can lead to an increase in the incidence of CRS. In this paper we use a mathematical model of the transmission dynamics of rubella virus to investigate the likely impact of different vaccination policies in Europe. The model was able to capture pre- and post-vaccination patterns of infection and prevalence of serological markers under a wide variety of scenarios, suggesting that the model structure and parameter estimates were appropriate. Analytical and numerical results suggest that endemic circulation of rubella is unlikely in Finland, the United Kingdom, The Netherlands, and perhaps Denmark, provided vaccine coverage is uniform across geographical and social groups. In Italy and Germany vaccine coverage in infancy has not been sufficient to interrupt rubella transmission, and continued epidemics of CRS seem probable. It seems unlikely that the immunization programmes in these countries are doing more harm than good, but this may be partly as a result of selective immunization of schoolgirls. Indeed, in both these countries, selective vaccination of schoolgirls with inadequate vaccination histories is likely to be an important mechanism by which CRS incidence is suppressed (unlike the other countries, which have had sufficiently high infant coverage rates to withdraw this option). Reducing inequalities in the uptake of rubella vaccine may bring greater health benefits than increasing the mean level of coverage.