Two fundamentally different approaches to vaccination of BALB/c mice with the hemagglutinin (HA) of A/Equine/Kentucky/1/81 (H3N8) (Eq/KY) were evaluated, that is, administration of HA protein vs administration of HA-encoding DNA. Each vaccine was tested for its immunogenicity and ability to provide protection from homologous virus challenge. HA protein was synthesized in vitro by infection of Sf21 insect cells with a recombinant baculovirus. Intranasal administration of this vaccine induced virus-specific antibodies, as measured by enzyme-linked immunosorbent assay (ELISA), but did not induce virus neutralizing (VN) antibodies. This route of administration provided partial protection from virus challenge, but interestingly, this protection was completely abrogated, rather than enhanced, by co-administration of 10 micrograms of cholera holotoxin. As a second approach, mice were directly vaccinated in vivo by Accell gene gun delivery of plasmid DNA encoding the Eq/KY HA gene. This approach induced VN antibodies as well as virus-specific ELISA antibodies. When two doses of DNA vaccine were administered 3 weeks apart, mice were not protected from challenge, although they cleared the infection more rapidly than control mice. However, when the second DNA vaccination was delayed until 9 weeks after the first, 9 out of 10 vaccinated mice were completely protected. These results indicate that the time between initial and booster DNA vaccinations may be an important variable in determining DNA vaccination efficacy.