A recent clinical sepsis trial reported a significant reduction in 90-day mortality by antithrombin (AT) exclusively in the subgroup of patients without simultaneous heparin prophylaxis. Patients additionally receiving heparin did not benefit from AT treatment. Herein, we studied the microhemodynamic and cellular mechanisms of this adverse effect of heparin on AT actions by the use of intravital microscopy and granulocyte culturing. In Syrian golden hamsters normotensive endotoxemia was induced by 2 mg/kg endotoxin (LPS, E. coli) i.v. In a first group of animals, AT (AT, 250 IU/kg i.v., n = 6) was given 5 min before LPS administration. A second group of animals (Heparin + AT, n = 5) received AT (250 IU/kg i.v.) combined with unfractionated heparin (sodium heparin, 100 IU/kg/24 h, i.v.). Additional animals (LMWH + AT, n = 5) received AT (250 IU/kg i.v.) combined with LMWH (nadroparin 47.5 IU anti-Xa/kg, s.c., 2 h before LPS). LPS-treated animals, which received only saline, served as controls (control, n = 6). Using dorsal skinfold fold preparations, LPS-induced microvascular leukocyte-endothelial cell interaction (LE) and alteration of functional capillary density (FCD) were studied by intravital video fluorescence microscopy. In controls, LPS induced a massive increase in LE with a maximum at 8 h and an impressive decrease in FCD over a 24-hour period. Both LPS effects were effectively prevented by AT treatment (p < 0.01), whereas Heparin + AT and LMWH + AT animals showed microcirculatory alterations comparable to that in controls. In additional in vitro chemotaxis assays. AT blocked neutrophil chemotaxis, an effect reversed by both unfractionated heparin and LMWH. Thus, our study elucidates a relevant in vivo and in vitro unfractionated heparin and LMWH adverse effect in the microcirculatory actions of AT during endotoxemia. These results indicate that heparin should be avoided to permit AT to modulate LPS-induced inflammatory responses.