Background: In a recent multicenter randomized trial comparing unfractionated heparin (UFH) with low-molecular-weight heparin (dalteparin) for thromboprophylaxis in 3,746 critically ill patients, 17 patients (0.5%) developed heparin-induced thrombocytopenia (HIT) based on serotoninrelease assay-positive (SRA 1) status. A trend to a lower frequency of HIT with dalteparin vs UFH was observed in the intention-to-treat analysis (fi ve vs 12 patients, P = .14), which was statistically signifi cant (three vs 12 patients, P = .046) in a prespecifi ed per-protocol analysis that excluded patients with DVT at study entry. We sought to characterize HIT outcomes and to determine how dalteparin thromboprophylaxis may reduce HIT frequency in patients in the ICU. Methods: In 17 patients with HIT, we analyzed platelet counts and thrombotic events in relation to the study drug and other open-label heparin, to determine whether the study drug plausibly explained seroconversion to SRA 1 status and/or breakthrough of thrombocytopenia/thrombosis. We also compared antibody frequencies (dalteparin vs UFH) in 409 patients serologically investigated for HIT. Results: HIT-associated thrombosis occurred in 10 of 17 patients (58.8%) (8:1:1 venous:arterial:both). Dalteparin was associated with fewer study drug-attributable HIT-related events (P = .020), including less seroconversion (P = .058) and less breakthrough of thrombocytopenia/thrombosis (P = .032). Antiplatelet factor 4/heparin IgG antibodies by enzyme-linked immunosorbent assay were less frequent among patients receiving dalteparin vs UFH (13.5% vs 27.3%, P < .001). One patient with HIT-associated DVT died after UFH bolus (anaphylactoid reaction), whereas platelet counts recovered in two others with HIT-associated VTE despite continuation of therapeuticdose UFH. Conclusions: The lower risk of HIT in patients in the ICU receiving dalteparin appears related to both decreased antibody formation and decreased clinical breakthrough of HIT among patients forming antibodies.