Thrombus aspiration catheters are devices used to remove a blood clot from a vessel, usually prior to angioplasty or stent implantation. However, in vitro results showed that the use of different commercial devices could produce very different thrombus removals, suggesting a primary dependence on the distal tip configuration of the catheter. A computational methodology based on realistic catheter tip modeling was developed to investigate the factors affecting the thrombus suction. Two different designs were considered, either with a single central lumen or a combination of central and side holes. First, steady-state aspiration of distilled water from a reservoir was simulated and compared with experimental tests. Subsequently, the aspiration of a totally occlusive thrombus, modeled as a high viscous fluid, was simulated solving a complex two-phase (blood and thrombus) problem. In particular, the benefit of additional openings was investigated. Good matching between the steady-state experimental and numerically simulated hydraulic behaviors allowed a validation of the numerical models. Numerical results of thrombus aspiration showed that the catheter with central and side holes had a worse performance if compared with the single central lumen catheter. Indeed, the inlets in contact with both blood and thrombus preferentially aspirate blood due to its much lower viscosity. This effect hindered the aspiration of thrombus. The amount of aspirated thrombus highly depends on the complex, two-phase fluid dynamics occurring across the catheter tips. Results suggested that location of additional holes is crucial in the catheter aspiration performance.