The recent outbreak of avian origin H10N7 influenza among seals in Northern Europe, and two fatal human infections with an avian H10N8 virus in China, have demonstrated that H10 viruses can spread between mammals and cause severe disease in humans. To gain insight into the potential for H10 viruses to cross the species barrier and identify a candidate vaccine strain, we evaluated the in vitro and in vivo properties and antibody response in ferrets to 20 diverse H10 viruses. H10 virus infection of ferrets caused variable weight loss and all 20 viruses replicated throughout the respiratory tract; however, replication in the lungs was highly variable. In glycan-binding assays, the H10 viruses preferentially bound "avian-like" α2,3-linked sialic acids. Importantly, several isolates also displayed strong binding to long-chain "human-like" α2,6-linked sialic acids, and exhibited comparable or elevated neuraminidase activity relative to human H1N1, H2N2, and H3N2 viruses. In hemagglutination inhibition assays, 12 antisera cross-reacted with ≥ 14 of 20 H10 viruses, and 7 viruses induced neutralizing activity against ≥ 15 of the 20 viruses. By combining data on weight loss, viral replication, and the cross-reactive antibody response, we identified A/mallard/Portugal/79906/2009 (H10N7) as a suitable virus for vaccine development. Collectively, our findings suggest that H10 viruses may continue to sporadically infect humans and other mammals, underscoring the importance of developing an H10 vaccine for pandemic preparedness. Avian origin H10 influenza viruses sporadically infect humans and other mammals; however, little is known about viruses of this subtype. Thus, we characterized the biological properties of 20 H10 viruses in vitro and in ferrets. Infection caused mild to moderate weight loss (5-15%), with robust viral replication in the nasal tissues and variable replication in the lung. H10 viruses preferentially bind "avian-like" sialic acids, although several isolates also displayed binding to "human-like" sialic acid receptors. This is consistent with the ability of H10 viruses to cross the species barrier, and warrants selection of an H10 vaccine strain. By evaluating the cross-reactive antibody response to the H10 viruses, and combining this analysis with viral replication and weight loss findings, we identified A/mallard/Portugal/79906/2009 (H10N7) as a suitable H10 vaccine strain.