Sera from rabbits hyperimmunized with hemagglutinin (HA) subunits isolated from the A/Port Chalmers/73 (H3N2)strain of influenza virus showed great differences in their cross-reactions with different strains of influenza virus. In hemagglutination-inhibition tests, some sera reacted to about the same titer with A/Port Chalmers/73 and A/Hong Kong/68 viruses, suggesting that these two strains were very closely related. Other sera, which reacted to high titer with A/Port Chalmers/73 virus, had only a low titer with the Hong Kong/68 strain, suggesting that the two viruses were distantly related. Evidence suggested that these diverse cross-reactions were due to widely different ratios, in the different sera, of antibodies to the "common" and the "specific" antigenic determinants on the HA subunits. Thus, some rabbits gave a stronger response to the "common" determinants than to the "specific", whereas in others, the reverse seemed to be the case. Sera from human volunteers injected with A/Port Chalmers/73 inactivated or subunit influenza virus vaccines, or from people infected with Port Chalmers/73 virus, contained, in most cases, antibodies predominantly to the "common" antigenic determinants on the HA subunits. These sera reacted to higher titer with Hong Kong/68 virus than with the Port Chalmers/73 strain. Absorption of these sera with Hong Kong/68 virus totally removed all detectable antibody, suggesting that they contained no antibody to the "specific" determinants of Port Chalmers/73 HA. Paradoxically, absorption of the sera with Port Chalmers virus did not remove all antibodies, suggesting that the sera contained antibodies to the "specific" determinants on Hong Kong/68 HA.