HLH, HTSH, HCG and HFSH are glycopolypeptides with similar molecular weights and immunological properties. Production of antisera useful in specific radioimmunoassays has been difficult. Immunization of rabbits with each of these hormones homogenized in complete Freund's adjuvant is routinely followed by antiserum production. Titers rise progressively with repeated immunizations for ten weeks or more and then the rate of increase falls. Rabbits producing higher titers of antiserum after two immunizations generally continue to produce higher titers during additional immunizations, thus selection of animals early during immunization is possible. Affinity of the antisera against all four hormones generally also increases during the course of immunization. Several ways of estimating affinity of antisera are presented. Specificity of antisera is variable, particularly against HFSH. Review of published techniques and our own experience reveals that all specific anti HFSH has been obtained after two to four immunizations; those after five or more immunizations have shown complete cross reaction with HLH, HTSH, and HFSH. These observations may explain the failure of some investigators to obtain specific antisera. For HFSH antisera, affinity progressively increases, but specificity appears to decrease with frequency of immunization. For HTSH and HCG antisera, specificity is retained or increases with repeated immunizations, and specific high affinity antisera are more regularly produced. Antisera capable of use to produce specific radioimmunoassays need not necessarily contain only antibodies against the hormone to be measured. Thus antisera against LH might be present in an antiserum used to quantify HFSH, if the anti HLH did not bind HFSH in either radioactive or nonradioactive form. Specificities of antisera vary a great deal and consideration of the nature of cross reacting substances would be effective in producing specificity. Thus when HLH reacted completely and equally with HFSH in an HFSH assay, absorbtion would completely neutralize all anti HFSH antibodies. If HLH and HFSH had dose response curves with quite different slopes (and hence different affinities for the anti HFSH), absorbtion is possible. When two or more populations of antibodies against HFSH are demonstrable by the presence of biphasic dose response curves in the assay, and if one population is specific, absorbtion of the nonspecific population may be possible.