Vaccination against Aujeszky's disease (AD) does not lead to eradication of the disease. Nevertheless, to reduce the economic losses, vaccination against AD is widely practised in many countries. With regard to control of AD, a method to distinguish infected from vaccinated pigs is a prerequisite in those countries. This paper deals with such a method. The assay (gI-ELISA) makes use of 2 monoclonal antibodies directed against 2 different epitopes on glycoprotein I (gI) of AD virus (ADV), and detects antibodies against those 2 epitopes. The gI-ELISA is sensitive, specific and suitable for serological screening on a large scale. Five of the 14 tested ADV vaccines did not elicit antibodies to gI in pigs, in other words they were gI-negative. On the other hand, all field strains of ADV tested so far did induce antibodies to gI. Pigs that were first given a gI-negative vaccine and subsequently were infected with a field strain of ADV produced antibodies to gI. These antibodies appear into the blood approximately 2 weeks after infection, and probably persist for over a year. Field experiments have demonstrated that it is possible to identify infected pigs in vaccinated herds. There is strong evidence, that ADV has been eliminated in a large multiplyer herd by removal of gI-seropositive pigs. The results described above provide a sound basis to implement a combined vaccination-control programme for AD in the Netherlands.