Serum samples from 105 patients with suspected heparin-induced thrombocytopenia (HIT) were evaluated using the 14C-serotonin release assay (SRA), considered the "gold standard" for the diagnosis of HIT, and two enzyme-linked immunosorbent assays (ELISA) that measure anti-platelet factor (PF) 4/heparin antibodies to determine the performance characteristics of the newly available ELISA assays. Relative to the SRA, the sensitivity and specificity of the Asserachrom HPIA assay were 73% and 77%, respectively, in this population of patients. The sensitivity and specificity of the GTI-HAT assay were 60% and 93%, respectively. In serum negative by SRA, GTI-HAT and HPIA detected antibodies in 9% and 25%, respectively. Antibodies were detected by HPIA in 18% of the sera negative by both SRA and GTI-HAT. In a second study, samples evaluated from patients (n = 10) treated for established thrombosis with a low-molecular-weight heparin and who had no decrease in platelet counts, showed a weak antibody titer in 50% of the patients after 12 days of therapy by GTI-HAT, whereas the HPIA identified a strong antibody titer in 75% of the patients after 4 days. These data suggest that the currently available ELISA methods for the detection of anti-PF4/heparin antibodies offer a limited sensitivity and specificity in comparison to SRA. The ELISA assays on their own are thus of limited value for the laboratory diagnosis of HIT. SRA-positive sera do not always have positive antibody titers, and antibodies can be present in SRA-negative sera. Furthermore, these data show that ELISA methods differ in their relative sensitivities and specificities for the detection of anti-PF4/heparin antibodies.