Cardiac emboli in patients with atrial fibrillation are one of the major causes of ischemic stroke. Because the vast majority of these cardiac emboli descend from the left atrial appendage (LAA), the therapeutic strategy of percutaneous transcatheter occlusion of the LAA by means of a mechanical implant was a logical consequence. Three different devices have been developed and used in humans: PLAATO, WATCHMAN, and ACP. The PLAATO implant, a “soccer ball” shaped self-expanding nitinol cage has demonstrated stroke prevention capability in small, uncontrolled studies. The WATCHMAN implant, a “half rugby ball” shaped self-expanding nitinol cage, has demonstrated non inferiority compared to warfarin treatment in the randomized, controlled PROTECT AF trial. The ACP implant is also a self-expanding nitinol cage. However, its design is of fundamental difference compared to the other two implants. ACP consists of two parts connected by a short waist, which are an “ice-hockey puck” shaped body for the implant fixation in the appendage wall, and a flexible disc for sealing the appendage ostium. Due to the very flexible connecting waist between the two parts, ACP implant adapts itself to the LAA, which appears in significant individual structural variabilities. The currently unpublished initial experience with ACP in Europe is encouraging. Transcatheter occlusion of the LAA offers an appealing way to reduce the incidence of cardioembolic stroke in patients with atrial fibrillation. However, the concerns about procedural safety and the need for long-term follow up should be addressed before this potentially important technology is deployed widely. The transcatheter occlusion of left atrial appendage, therefore, has the potential to become, but is not yet a real alternative to anticoagulation.