In May 1993, a severe epidemic of respiratory disease began in horses in Inner Mongolia and spread throughout horses in China. The disease affected mules and donkeys as well as horses but did not spread to other species, including humans. The severity of the disease raised the question of whether the outbreak might have been caused by the new avian-like influenza viruses detected in horses in China in 1989 or by current variants ofA/equine/Miami/1/63 (H3N8) (equine-2) or by a reassortant between these viruses. Antigenic and sequence analysis established that all gene segments of the influenza virus causing the epidemic were of recent equine-2 origin and that the virus was not a reassortant. Serological analysis of post-infection horse sera provided evidence for the continued circulation of the A/Equine/Jilin/1/89 (Eq/Jilin) (H3N8) avian-like viruses in horses in Heilongjiang province with original antigenic sin-like responses. It is noteworthy that prior infection with the avian-like Eq/Jilin strain did not afford cross-protection against a current equine-2 strain. Serological evidence for the continued circulation of the avian-like H3N8 influenza virus in horses indicates that this virus has probably established itself in horses in Asia.