Parnaparin, a low-molecular-weight heparin (LMWH), prevents platelet activation and interaction with polymorphonuclear leukocyte (PMN) in a washed cell system. The in-vitro effect of parnaparin was studied here on platelet-PMN aggregates formed with more physiologic approaches in whole blood, in parallel with unfractionated heparin and enoxaparin, another LMWH. Citrated blood from healthy subjects was stimulated: i) from passage through the "Platelet Function Analyzer" (PFA-100), a device that exposes blood to standardized high shear flow through collagen/ADP cartridges; ii) by collagen and ADP (2 and 50 mug/ml, respectively) added in combination under stirring in an aggregometer cuvette; iii) with recombinant Tissue Factor, to generate thrombin concentrations able to activate platelets without inducing blood clotting, or iv) the Thrombin Receptor Activating Peptide-6 (TRAP-6). Platelet P-selectin and platelet-PMN aggregates were measured by flow cytometry upon stimulation of blood. Fibrinogen binding to platelets and markers of PMN activation were also detected. Platelet P-selectin expression and platelet-PMN aggregate formation were induced in all four activation conditions tested. Parnaparin prevented in a concentration-dependent manner (0.3-0.8 IUaXa/ml) the expression of P-selectin and the formation of mixed aggregates, while the two reference heparin preparations had a much weaker effect. Platelet fibrinogen binding and PMN activation markers (fibrinogen binding, CD11b and CD40) were also prevented by parnaparin. These data extend in more physiological systems of platelet activation, the anti-inflammatory profile of parnaparin, previously reported in washed cells. The greater effect of parnaparin, as compared to the reference heparins, could be due to chemico-physical differences possibly unrelated to their anticoagulant effect.