Sera from volunteers who received live influenza A wild-type or ts recombinant virus were tested by hemagglutination inhibition (HI) assay, neuraminidase inhibition (NI) assay, and the enzyme-linked immunosorbent assay (ELISA) to determine which assay system was the most sensitive in detecting an immunological response to infection. The ELISA was performed with inactivated whole virus antigen, and the optical density at each of five serial twofold dilutions of pre- and postimmunization sera was measured. The difference in the amount of ELISA antibody in pre- and postinoculation serum specimens was taken to be proportional to the area between the respective titration curves. The ELISA was more sensitive than the HI or NI test in detecting a seroresponse in volunteers infected with A/Hong Kong/123/77 (H1N1), A/New Jersey/8/76 (Hswine N1), or A/Alaska/6/77 (H3N2) ts recombinant virus. These results suggest that the ELISA should be used to determine the frequency of infection with attenuated viruses as well as the 50% human infectious dose of candidate live influenza A vaccine viruses.