The serum antibody response of infants and children immunized with Formalin-inactivated respiratory syncytial virus (RSV) vaccine 20 years ago was determined by using an enzyme-linked immunosorbent assay specific for the RSV fusion (F) and large (G) glycoproteins and a neutralization assay. Twenty-one young infants (2 to 6 months of age) developed a high titer of antibodies to the F glycoprotein but had a poor response to the G glycoprotein. Fifteen older individuals (7 to 40 months of age) developed titers of F and G antibodies comparable to those in children who were infected with RSV. However, both immunized infants and children developed a lower level of neutralizing antibodies than did individuals of comparable age with natural RSV infections. Thus, the treatment of RSV with Formalin appears to have altered the epitopes of the F or G glycoproteins or both that stimulate neutralizing antibodies, with the result that the immune response consisted largely of "nonfunctional" (i.e., nonneutralizing) antibodies. Subsequent natural infection of the vaccinees with wild-type RSV resulted in enhanced pulmonary disease. Despite this potentiation of illness, the infected vaccinees developed relatively poor G, F, and neutralizing antibody responses. Any or all of three factors may have contributed to the enhancement of disease in the RSV-infected vaccinees. First, nonfunctional antibodies induced by the inactivated RSV vaccine may have participated in a pulmonary Arthus reaction during RSV infection. Second, the poor antibody response of infants to the G glycoprotein present in the Formalin-inactivated vaccine may have been inadequate to provide effective resistance to subsequent wild-type virus infection. Third, the relatively reduced neutralizing antibody response of the infant vaccinees to wild-type RSV infection may have contributed to their enhanced disease by delaying the clearance of virus from their lungs.