Serum samples from 156 Greek persons were assessed by an IgG-specific enzyme-linked immunosorbent assay (ELISA) and a reference tissue culture toxin-neutralization (TN) assay for the quantitation of diphtheria toxin antibodies. By the reference method, 7.7% of the persons were susceptible to diphtheria (antitoxin < 0.01 IU/ml), 28.8% had basic protection (antitoxin 0.01-0.09 IU/ml) and 63.5% were fully protective (antitoxin > or = 0.1 IU/ ml), while the corresponding figures were 17.9, 36.5 and 45.5% when they were tested by the immunoassay. None of the samples been susceptible by the TN assay were found to have some protection when tested by ELISA. However, three (6.7%) of the 45 samples showing a basic protection with TN, were fully protective when titrated by the immunoassay. In addition, 31 (31.3%) of the 99 samples been fully protective by the bioassay, were found to be either basically protective or susceptible by means of the ELISA. Overall, validity features of the immunoassay were: sensitivity 68.7%, specificity 94.7%, positive predictive value 95.8% and negative predictive value 63.5%. The ELISA tested in our study could be used to determine diphtheria antitoxin in individuals needed a booster immunization (susceptible or basic protective samples), although it might falsely include in the above categories samples that are within the fully protective levels of antibodies.