Groups of 60 to 120 mice were given a single intraperitoneal inoculation of varying dilutions of commercially prepared and licensed bivalent (A/England and B/Mass) and monovalent (A/Aichi or B/Hong Kong) inactivated influenza vaccines, and their antibody responses at 14 days were quantitated by hemagglutination inhibition tests. Split-product vaccines prepared by the treatment of A/England, B/Mass, and B/Hong Kong whole virus with Tween-80 and either tributylphosphate or ether produced significantly lower mean antibody titers than did equivalent whole-virus preparations. The rates of seroconversion (<1:8 to >/=1:8) at the various dilutions tested were also significantly reduced when these split-product vaccines were given. When the antigen content of all vaccines was quantitated by the chick cell agglutination test, between 10 and 100 times more split-product antigen than whole-virus antigen was required to produce seroconversion in 50% of the mice tested. Differences between split-product and whole-virus A/Aichi vaccines were less marked. These data point out the need to consider factors other than hemagglutinin content alone in determining the immunogenicity of inactivated influenza vaccines.