Recent studies have demonstrated a strong association between type II (immunologically mediated) heparin-induced thrombocytopenla/thrombosis (HITP) and antibodies reactive with complexes consisting of heparin and platelet factor 4 (PF4), a heparin-binding protein normally found in platelet-or granules. However, the frequency with which such antibodies develop in patients given treatment with heparin has not yet been defined. We studied the development of heparin:Pf4-specific antibodies in 51 patients who received a single dose of unfractionated heparin (UFH) during cardiac catheterization and were then given UFH or low-molecular-weight heparin (LMWH) again during and after open heart surgery. Eleven of the 51 patients (22%) had antibodies reactive with heparin:PN when they were admitted for cardiac surgery; these antibodies were mainly of the Immunoglobulin M (IgM) class and were apparently stimulated by exposure to UFH at cardiac catheterization. Seventeen of 34 patients (50%) without preexisting antibody who were given UFH during and for 1 to 3 days offer surgery formed immunoglobulin G antibodies or IgM antibodies (or both) by the sixth postoperative day. Overall, 27 of 44 patients (61%) who were given UFH at surgery had antibodies by the time of hospital discharge. None of 6 patients without preexisting antibody who were given LMWH at surgery formed antibodies (p < 0.03). However, LMWH was given as a single injection only on the day of surgery. The titer of the antibodies formed by patients receiving UFH ranged from 1:10 to 1:200, significantly lower than those in patients with a clinical diagnosis of HITP. Moderate thrombocytopenia was common after open heart surgery, but platelet levels in patients who had preexisting antibodies or formed new antibodies did not differ significantly from those in patients without antibody. Clinically significant thrombosis did not develop in any patient and HITP was not diagnosed in any patient. Antibodies reactive with heparin:PF4 formed in only 3 of 66 patients (4.5%) undergoing other types of surgery, One of these patients had been given UFH 3 months previously; the other 2 may have been exposed to heparin used to flush intravenous lines postoperatively, No antibodies reactive with heparin:PF4 were found in any of 108 normal subjects. We conclude that UFH is more immunogenic than has been thought and that patients exposed to this anticoagulant during open heart surgery are at high risk to form low titer (less than or equal to 1:200) antibodies reactive with heparin:PM. Further studies are needed to determine whether such antibodies are clinically significant-that is, whether sensitized patients are at risk to develop HITP if heparin treatment is continued for more than I to 3 days or is reinstituted at a later date.