Influenza is an acute respiratory disease caused by influenza type A and B viruses. Human influenza viruses may infect up to 15% of the total population during the seasonal epidemics, causing many cases of severe illness. Each year, approximately 350 million doses of influenza vaccine are produced for protection of those at risk of severe disease. The thesis focuses on the protection against influenza virus infection and disease (Paper I and II) as well as the analysis of the antigen variations found in primarily the hemagglutinin (HA) gene of the virus (Paper III and IV). The time required for the production of influenza vaccines is 6-8 months. A more effective and rapid method of production is desirable, both for the annual epidemics and in case of an influenza pandemic. A DNA-based vaccine could be a useful alternative. In paper I, the immunological response in ferrets after intramuscular immunisation with a plasmid construct expressing chimeric influenza HA proteins was evaluated. Strain specific antibodies were elicited, but none of the ferrets immunised with a DNA or subunit vaccine were protected from infection when challenged with an influenza A/H3N2 virus homologous or heterologous to the vaccine. Considerable enhancement of the immune response induced by DNA immunization will be needed before the approach can be a realistic alternative for vaccination of humans. Hem agglutination inhibition (HAI) is the standard method for determination of protective levels of influenza antibodies. The method is not efficient for all virus subtypes and strains, and alternative methods suitable for large-scale examinations are desirable. In paper II, an in situ neutralisation test (NT) for the measurement of influenza antibodies was created and evaluated in two human cell-lines, human fibroblasts (HS27) cells and human salivary gland epithelial duct (HSG) cells, and in Madin-Darby canine kidney (MDCK) cells. The HS27 cell line gave stable results and was most suitable for antigen detection with enzyme-linked ...