Two obstacles hinder the development of an AIDS vaccine: (1) the AIDS virus exhibits extensive amino acid heterogeneity between isolates and (2) antibodies elicited by virus during the course of natural infection are often non-neutralizing. A vaccine designed to induce anti-idiotypic antibodies against the virus' receptor on T-cells, T4, should, in principle, overcome these obstacles. Such antibody could contain an "internal image" of T4 and bind the receptor binding domain of the virus. Since this domain is both critical to function and, therefore, poorly susceptible to antigenic variation, anti-receptor anti-idiotypic antibodies may demonstrate broad, strain-independent crossreactivity and block viral adherence.