The first cause of congenital heart disease is atrial septal defect (ASD). Several devices were developed to close secundum ASD but complications can occur (thromboembolic, infectious…) [1] , [2] , [3] . Providers claim different coatings supposed to improve device haemocompatibility. To study, in vitro, the ability of 3 devices to be endothelialized by endothelial cells derived from circulating human endothelial progenitors cells (EPCs), to induce platelet or complement activations or coagulation intrinsic pathway activation. EPCs from umbilical cord blood were extracted, cultured and characterized (CD31, VE-cadherin and von Willebrand factor). Devices were seeded with 100,000 cells/cm 2 . EPC adhesion, at 3 and 24 hours, was investigated (biological activity of N acetyl β-D-hexosaminidase). EPC proliferation was monitored with Alamar blue ® test which allowed a longitudinal follow-up (Days 1, 3, 6, 8, 10 and 12). C3a assay was performed after blood-contacting devices (standard ASTM F1984) [4] . Thereafter, platelet activation (via Pselectin and GPIIBIIA) and blood coagulation on biomaterials (standard ASTM F2888, F2382) [5] , [6] were explored. With regard to EPCs adhesion and proliferation ( Fig. 1 , Fig. 2 ), no statistically significant differences were found between 3 devices. There was a significant EPC proliferation on each device as a function of time appearing at Day 8 for devices 2 and 3 and Day 10 for device 1. No complement activation was detected by the C3a assay (device 1: 3584 ± 978 ng/mL, device 2: 3386 ± 1092 ng/mL and device 3: 4612 ± 1657 ng/mL) ( Fig. 3 ). No platelet activation occurred within 15 min of contact with devices (Figs 4 and 5). However, there was a minimal activation of coagulation for 3 devices (Mean time for PTT test on device 1: 183 ± 23.3 s, device 2: 170.7 ± 22.2 s and device 3: 167 ± 22.6 s). Despite different coatings, the haemocompatibility of three devices was comparable. The benefit of effective anticoagulation should be confirmed by future clinical studies.