MPEG is traditionally committed to coding and compression of audio-visual data from natural sources. However, the emerging MPEG-4 standard aims not only at multiple natural audio-visual objects composing the scene, but also synthetic audio and video to be integrated with the natural. It will also allow more interaction with both synthetic and natural objects. Networked Collaborative Virtual Environments (NCVE) have a wide range of different requirements on networking. This paper analyzes how these requirements could potentially be met by MPEG-4 and shows that MPEG-4 tools could be extremely useful for development of such environments.