Inactivated porcine parvovirus vaccines have been available commercially in Britain since 1984 and are now widely used in breeding herds. To investigate their value in cost benefit terms an oil-emulsion vaccine developed at Weybridge was used in trials on 1243 gilts in 12 herds during the period 1984 to 1986. In each herd approximately half the gilts were given the vaccine before breeding and the remainder were left unvaccinated. Blood samples were taken at vaccination and two to four weeks later to measure the serological responses, and the reproductive performances of the two groups were compared. When the data from all the gilts in the 12 herds were combined and analysed together there was surprisingly little difference between the reproductive performance of the vaccinated and unvaccinated groups. Only when the results from individual herds were analysed and interpreted against a background knowledge of wild parvovirus activity (as derived from a study of the serological results) did an understanding and evaluation of the benefits of vaccination become possible. As herds vary with respect to the absence or presence of porcine parvovirus and the epidemiology of the infection it is recommended that vaccination be used with discrimination; it should then prove highly cost effective.