Introduction: A dynamic model to evaluate thrombus formation on intravascular catheters in vitro is presented. The model enables fluid infusion, variation in the catheter orientation, and variable flow conditions. It was applied on a catheter used to shunt cerebrospinal fluid to a vein, a dural venous sinus, for the treatment of hydrocephalus. Methods: Fresh human blood-filled circuits were circulated in a non-occlusive roller pump. A catheter infused either with cerebrospinal fluid, Ringer's lactate, or no fluid (control) was inserted through each circuit's wall. Sixteen circuits (six cerebrospinal fluid, six Ringer's lactate, four control) ran for 60 min. Qualitative assessment was performed by measuring viscoelastic properties of blood at the start and end of the experiment; quantitative evaluation of clot formation by scanning electron microscope. Results: Average blood velocity was 79 mm/s, with a pressure wave between 5 and 15 mm Hg. At the experiment's end, the infused fluid represented 5.88% of the blood/infusion volume in the circuit. The control circuits showed no statistical difference between the start and end for viscoelastic testing, whereas both Ringer's lactate and cerebrospinal fluid enhanced coagulation, most pronounced for the latter. Most thrombus material was observed on catheters in the cerebrospinal fluid group. Clot formation was less pronounced on the surface of the catheter facing the blood flow. Discussion: A dynamic model for intravascular catheter testing mimics better clinical conditions when evaluating blood-material interaction. Catheter position, blood flow around the catheter, and infusion fluid all have a potential impact on the hemocompatibility of a given catheter.