We performed a retrospective study on the prevalence of heparin-induced thrombocytopenia (HIT) in 233 patients receiving hog mucosa heparin therapy. Of these, 82 patients received s.c. calcium heparin, 130 patient received unfractionated (UF) i.v. heparin, and 21 patients received low molecular weight heparins (LMWH). An additional four patients, referred to our consultation and diagnosed by us as having clinically active type II HIT (HIT-II) were also studied. The mean platelet count of the 233 patients receiving heparin showed a significant decrease after 2 days of heparin treatment and a following significant increase 6 days later (basal: 257 ± 147 x 10 9 platelets/L; day 2: 239 ± 122, p < 0.0002; day 6: 286 ± 119, p < 0.004). Of the 212 patients receiving UF heparin, 13 (6%) fulfilled the criteria for HIT-II: seven of these had received i.v. heparin (mean daily dose 26,600 ± 4,082 IU & p l u s m n ; SD) and six had received s.c. heparin (mean daily dose 21,428:t 6,900 IU). Their mean basal platelet count was 226 & p l u s m n ; 100 SD × 10 9 platelets/L and the nadir during heparin treatment was 78 ± 39 x 10 9 platelets/L. Thrombotic complications occurred in four (30.7%) of the 13 patients with HIT-II. Since the immunological mechanism has been demonstrated for HIT-II and since platelet factor 4 (PF4) was identified as the co-factor for the binding of heparin-related antibodies, we set up our own enzyme-linked immunosorbent assay (ELISA) for testing antibodies against PF4/heparin complex bound through electrostatic bridges to the solid phase. The highest binding capacity of HIT-related IgG to the multimolecular complex was obtained at 20 μg/ml for PF4 and 3 μg/ml for heparin, corresponding to 250 ng of PF4 and 42 ng of heparin in each microtiter well. Such binding was inhibited in a dose-dependent manner by increasing amounts of heparin, protamine hydrochloride, and a monoclonal antibody anti-human PF4 clone 1OB2. We observed that HIT-related antibodies bound also to PF4/LMWH complexes but the optimal PF4/glycosaminoglycan ratio appeared more critical for LMWH (enoxaparin, fraxiparin, and pamaparin) than for UF heparin. Sera from eight patients with HIT-II were tested by PF4/heparin ELISA; six of these had IgG against the complex PF4/heparin and three also had IgM. The persistence of HIT-related antibodies was investigated in three patients: in one such antibodies were still detectable 3 years after the acute episode, while in the other two, they disappeared after 6 months and 1 year, respectively.