Abstract: In this work, we propose an innovative approach, 3D recording, to support serious games and educational role-playing. 3D virtual worlds (3D VWs) are often used for generating educational content. Even though this technology allows creating full context of the real-life educational process, it is usually recorded as flat 2D video (such as Machinima in Second Life), which eliminates many advantages of the technology, such as sense of presence (Mckerlich et al. 2011). In addition, there are no systematic approaches for combining synchronous and asynchronous learning modes. We propose that 3D recording is capable of solving these challenges, as it offers an easy way for creating advanced 3D content from synchronous activities, as well as accessing this content in both synchronous and asynchronous modes.