Equine influenza viruses of the H3N8 subtype, which have circulated in horses in various parts of the world, have since 1989 comprised two lineages, distinguishable by HA sequence and HI tests using post-infection ferret antisera: the 'American' lineage, represented by A/eq/Newmarket/1/93 or A/eq/South Africa/4/2003, and the 'European' lineage, represented by A/eq/Newmarket/2/93.

Since 2003, few European-lineage viruses have been isolated and none were isolated in 2007. Two principal clades (based on HA sequence) of American-lineage viruses have circulated in recent years, clade 1 viruses predominantly in the USA, Japan and in Australia for the first time in 2007, and clade 2 viruses mainly in Europe.

Equine influenza in 2007 was characterized by major epizootics in Japan (the first since 1972) and Australia (the first detection of equine influenza) beginning in August and extensive outbreaks in China and Mongolia during October and November, as well as reports of widespread disease in Kazakhstan and Sweden and numerous small outbreaks in Europe and the USA.

The viruses responsible for the Australian and Japanese epizootics, e.g. A/eq/Sydney/2007, were closely related to A/eq/South Africa/4/2003, the prototype American-lineage vaccine strain recommended since 2004. (Animal Health Trust).

The most recent recommendation on vaccine composition (2013) for equine influenza is that it is not necessary to include an H7N7 virus or an H3N8 virus of the Eurasian lineage in vaccines as these viruses have not been detected in the course of recent surveillance and are therefore presumed not to be circulating.

Vaccines for the international market should contain both clade 1 and clade 2 viruses of the Florida sublineage. Clade 1 should be represented by A/eq/South Africa/04/2003-like or A/eq/Ohio/2003-like viruses and Clade 2 should be represented by A/eq/Richmond/1/2007-like viruses.