This study describes a modification of Vane's blood-bathed organ technique (BBOT). This new technique consisted of replacing the cascade of contractile smooth muscle organs within the traditional BBOT by a single collagen strip cut from a rabbit's hind leg tendon. Utilizing the extracorporeal circulation of an anesthetized heparinized mongrel cat or Wistar rat, arterial blood was dripped (1-3 ml min(-1)) over a collagen strip. This resulted in a gain in weight of the strip, which was due to the deposition of platelet aggregates and a few blood cells trapped over the strip. Arterial blood that had been used for the superfusion was pumped back into the animal's venous system. However, when this technique is adapted to human volunteers, the superfusing blood should be discarded. In animal experiments, intravenous injections of a variety of classic fibrinolytic agents (e.g., streptokinase) promoted the formation of platelet thrombi. Nitric oxide donors (e.g., SIN-1) at non-hypotensive doses hardly affected the mass of platelet thrombi deposited over the collagen strip, whereas endogenous prostacyclin (e.g., released from vascular endothelium by bradykinin) or exogenous prostacyclin and its stable analogues (e.g., iloprost) dissipated platelet thrombi as measured by a loss in the weight of the blood superfused collagen strip. This model allowed us to assay numerous drugs for their releasing properties of endogenous prostacyclin from vascular endothelium. These drugs included lipophilic angiotensin converting enzyme inhibitors (ACE-Is), which act in vivo as bradykinin potentiating factors (BPF). Other PGI(2)-releasers included statins (e.g., atorvastatin and simvastatin), thienopyridines (e.g., ticlopidine and clopidogrel), a number of thromboxane synthase inhibitors, flavonoids, bradykinin itself, cholinergic M receptor agonists and nicotinic acid derivatives. The thrombolytic actions of lipophilic ACE-Is (e.g., quinapril and perindopril) were prevented by pretreatment with either bradykinin B(2) receptor antagonists (e.g., icatibant) or with endothelial COX-2 inhibitors (e.g., rofecoxib, celecoxib and high dose aspirin). The inhibition of endothelial nitric oxide synthetase (eNOS) by L-NAME hardly blunted the thrombolytic response to ACE-Is. Hence, it can be concluded that many recognized cardiovascular drugs apart from their known basic mechanisms of action, may also behave as releasers of endogenous endothelial prostacyclin. Furthermore, in many instances, this effect may be the primary mechanism of their therapeutic efficacy.