The enzyme-linked immunosorbent (ELISPOT) assay, which enumerates peripheral blood mononuclear cells (PBMCs) releasing interferon gamma (IFN-gamma) on specific antigen stimulation, is becoming the assay of choice for evaluation of vaccine-induced cell-mediated immune responses in many clinical trials. A properly conducted trial requires the assays to be validated, especially should the trial lead to vaccine licensure. Here, the design and validation of an ELISPOT assay are described for use in clinical trials of candidate human immunodeficiency virus (HIV) vaccines, using a particular immunogen termed HIVA. This assay employs eight pools of 20 to 23 peptides each: seven pools are derived from the immunogen and one pool is derived from cytotoxic T cell epitopes of common human viruses serving as an internal positive control. The validation determined that first, the overall variation of a positive response of approximately 500 spot-forming units (SFU)/10(6) cells was 21%, while second, the average of 5 SFU/10(6) cells was detected for the seven HIVA-derived pools in HIV-uninfected individuals; third, a positive response to a peptide added to the assay pools was not occluded by the other pool peptides; fourth, the frequencies detected in fresh PBMCs were 2- to 3-fold higher compared with the same samples that had been cryopreserved; and finally, all seven HIV-derived pools induced IFN-gamma responses in PBMCs isolated from HIV-infected individuals. The limits of the validation of assays involving biological responses of living cells are discussed.