Antigenic analysis of human and avian H2 influenza viruses were done with monoclonal antibodies to the HA molecules in hemagglutination inhibition (HI) assays. These studies revealed that the receptor-binding specificity of the hemagglutinin can markedly influence the antigenic analysis obtained with monoclonal antibodies in HI tests. Influenza viruses that are sensitive or resistant to inhibition by horse serum inhibitors showed marked differences in their reactivity with monoclonal antibodies to the hemagglutinin. This was apparent with the A/RI/5+/57 and A/RI/5-/57 strains of H2N2 viruses isolated by Choppin and Tamm (1960a), half of the panel of different monoclonal antibodies failed to inhibit hemagglutination of the RI/5- variant, whereas all of the 18 monoclonal antibodies inhibited RI/5+. These findings have important implications in the antigenic analysis of influenza viruses where HI assays are conventionally used to determine the extent of antigenic drift in nature. Antigenic differences were detectable between different human H2 influenza virus isolates from 1957 that were sensitive to inhibition by horse serum, indicating that minor antigenic variation occurs within the first year of appearance of the new subtype. Minor antigenic variation continued in the H2 viruses until 1961, but by 1962 antigenically distinguishable variants that could be discriminated with both monoclonal antibodies and postinfection ferret antisera predominated. Analysis of avian H2 influenza viruses with a panel of monoclonal antibodies indicated that antigenic variation occurs and that multiple different variants cocirculate in the population. There was no progressive antigenic change in the avian H2 influenza viruses with time, as was found with the human H2N2 strains. Topographical mapping of the H2 hemagglutinin by selection of antigenic variants with monoclonal antibodies and analysis of their reactivity patterns by HI showed overlap between the epitopes examined. These results may reflect restriction in the antibody repertoire of the mice used in preparation of the monoclonal antibodies or that the H2 hemagglutinin does not have such discrete nonoverlapping antigenic regions found in the early H3 influenza virus.