Aim. Influenza A is a virus within the Orthomyxoviridae family causing an acute contagious viral respiratory infection and circulating among people worldwide. In humans, influenza A viruses cause yearly outbreaks with high morbidity and excess fatality rates. Influenza A natural hosts are the dabbling ducks, where it occurs as a benign disease. However, it can infect and adapt to various hosts. It has error-prone replication strategies that allow to undergo rapid genetic change by recombining with other influenza viruses to form viral subtypes. The adaptation of the virus depends on the stability of hemagglutinin (HA) protein, balance of HA binding, neuraminidase (NA) receptor-destroying activities, receptor-binding specificity by the HA protein, HA stability, polymerase (PA) efficiency etc. People are more often infected by IV subtypes specific to human, and birds – by the avian specific subtypes because of different combination and location of receptors in respiratory system. Pigs and ferrets can be infected by both viral subtypes. Because of interaction of different subtypes, these species are known to be potential sources of new variants of IV with ability to cross species transmission. The aim of the study was testing if the positive selection drives the evolution of HA\NA\PA by favoring mutations that enhance viral fitness, such as by increased binding affinity to host receptors or evasion of host immune responses. We also tested if the swine are the hosts in which virus selection is intensified. Conclusions. As expected, the chosen proteins are under positive selection. We found the evidence of virus positive selection in the PA samples from New York (27 sites), HA and NA samples from Egypt (17 and 16 sites, respectively). The most often mutations for HA were found in mature peptide HA1 domain, for NA – in sialidase β-propeller 6 structural motif, and PA – in Arch domain. No evidence of positive selection for all 3 proteins for the samples from Bangladesh, that can be a signal of slow evolutionary activity at this country. The testing of evolution in swine hosts showed that selection is significantly relaxed in the test group. Which means that genetic drift, but not a selection is a main force in the test group. The result might be different in datasets with more swine samples.