Purpose: To evaluate if CFD can determine the risk of delayed rupture of aneurysms treated with flow diverters. Materials and Methods: Two carotid aneurysms each treated by a single SILK were studied. One completely occluded within 6 months. CTAngiogram before stenting, at 2 days (40% thrombosed) and 6 months were analysed. The other ruptured at 3.5 months. The 3 months CTA (no aneurysm thrombosis) was studied. CFD performed using ANSYS. The Virtual FD was simulated using the porous media method. Results: For the occluded aneurysm, CFD of the first CTA after placement of a Virtual FD showed a reduction of energy loss across the aneurysm neck of 64%, reduction of flow into the aneurysm by 92.9%, disappearance of the high wall shear stress (WSS) region and change in the flow pattern from a circular jet coming from the inflow to a non-jet flow from the FD centre with stasis in large part of the aneurysm. This predicted the actual day 2 results showing further reduction to 16% and by 99.9% respectively, uniform low WSS and minimal central flow only. Actual energy loss at 6 months was only 3%. For the ruptured case, the Virtual FDs simulated the actually deployed Silk which was stretched at the inlet allowing a large jet to flow into the aneurysm and simulated an optimally deployed FD. The simulated stretched versus optimal FD studies showed energy loss of 63% versus 52%, flow reduction of 43.8% to 96.3%, continued jet flow to no jet flow, and high WSS to low WSS respectively. Conclusion: The CFD findings of high energy loss, low flow reduction, continued jet flow and high WSS are possible predictors of delayed rupture. Virtual FD can be used to predict outcome.