OBJECTIVE A few case reports have suggested a possible thrombogenic effect of nonionic contrast media. In vitro investigations have lead to conflicting results. The authors performed three ex vivo studies to evaluate the influence of an ionic, ioxaglate, and a nonionic, iopamidol, low-osmolality contrast medium on a series of clotting and fibrinolytic parameters, after intravenous or intra-arterial administration, during routine diagnostic procedures. METHODS In the first study, iopamidol was given to 20 consecutive patients through an arterial catheter for digital subtraction arteriography (DSA). In the second study, iopamidol was compared with ioxaglate. The media were randomly and blindly administered intravenously to 21 consecutive patients undergoing brain computed tomography (CT). Finally, ioxaglate was administered intra-arterially to 20 consecutive patients, in a situation comparable with that of the first study. RESULTS In the first study, a weak anticoagulant effect and an activation of fibrinolysis were found, associated with indirect markers of thrombin generation, such as increased plasma levels of fibrinopeptide A (FpA) and thrombin-antithrombin III complexes (TAT). In the second study, no significant changes were seen with either contrast medium, for thrombin or fibrinolysis activation parameters. In the third study, the intra-arterially administered contrast medium elicited a marked increase of FpA and TAT, together with an anticoagulant effect. CONCLUSIONS Both ionic and nonionic contrast media are able to interfere with the clotting/fibrinolytic system in the general circulation when they are administered to patients at the usual dosages. Ioxaglate shows more marked anticoagulant and thrombin-generating effects than iopamidol. The procedure (ie, arterial catheter versus intravenous infusion) seems to be more important than the category of contrast medium in conditioning the magnitude of these effects.