Since results of recent clinical trial, performed in patients suffering from acute coronary syndromes, suggested that aspirin improved the efficacy of an anti-GP-IIb/IIIa therapy, this work was set to study the mechanism underlying such a mechanism of aspirin. SR121566 (3 mg/kg, i.v.), a new selective antagonist of the GP-IIb/IIIa complex inhibited thrombus formation which occurred following electrical stimulation of the carotid artery of rabbits. On rabbits pre-treated with aspirin (10 mg/kg, i.v.), a bolus injection of SR121566 produced a strong reversal of thrombosis compared with the effect obtained on rabbits treated with SR121566 alone. In vitro, the addition of SR121566 to human platelets during the ADP-induced aggregation process resulted in a reversal of the aggregation which was nearly complete when it was added 30 s after the induction of aggregation, but progressively decreased when SR121566 was added at later times. ADP-induced platelet aggregation in the presence of aspirin (1 mM) was comparable to that obtained in control platelet-rich plasma, but the addition of SR121566 at all time points of the aggregation process was followed by a complete reversal of the aggregation even at later time points (> 120 s). Histological examination of the platelet aggregates showed a clear-cut difference in the platelet content between control and aspirin-treated platelets which contained much more dense granules. Inhibition of the release of thrombospondin by aspirin may account for a significant part of its pro-desaggregating activity as demonstrated by the effect of the tetrapeptide VTCG, an inhibitor of the binding of thrombospondin to its receptor, which strongly potentiated the activity of SR121566 as a desaggregating agent. Therefore, our study shows that aspirin pretreatment promotes a 'thrombolytic' activity of GP-IIb/IIIa inhibitors and, in particular, of SR121566, a potent and selective member of this new class of compounds. This effect is mainly due to an inhibition of the release of the platelet granule content and in particular to thrombospondin. This observation may contribute to a greater efficacy of GP-IIb/IIIa inhibitors for the treatment of thrombosis in atherosclerotic patients.