The present paper reports on the development, validation and field application of a control strategy for avian influenza infections in poultry. The "DIVA" (Differentiating Infected from Vaccinated Animals) strategy is based on the use of an inactivated oil emulsion vaccine containing the same haemagglutinin (H) subtype as the challenge virus, but a different neuraminidase (N). The possibility of using the heterologous N subtype, to differentiate between vaccinated and naturally infected birds, was investigated through the development of an "ad hoc" serological test based on the detection of specific anti-N antibodies. This test is based on an indirect fluorescent antibody assay, using as an antigen a baculovirus expressing recombinant N proteins. The vaccination strategy has been tested in the laboratory and shown to be efficacious both against challenge with highly pathogenic AI viruses and with low pathogenicity AI viruses, ensuring clinical protection, reduction of duration and titre of shedding. In addition, vaccination resulted in an increased resistance to infection. The companion diagnostic tests directed to the detection of anti-N1 and anti-N3 antibodies have been validated in the laboratory and using field samples. The serological assay showed an "almost perfect agreement" (Kappa value) with the HI test, with relative sensitivity and specificity values of 98.1 and 95.7, respectively. The results of the present investigation suggest that the "DIVA" control strategy may represent a tool to support the eradication of avian influenza infections in poultry.