A multimedia on demand system includes a multimedia server configured to connect to an internet-based network and to transmit audio/video programs on demand via said internet-based network. A plurality of customer premises units connected to the internet-based network via a network interface are configured to receive the audio/video programs via the internet-based network. A remote control unit in communication with each customer premises unit provides user input to the unit. An audio video system is connected to each customer premises unit for reproduction of the audio/video programs. The system can be used to provide multimedia on demand services over the internet, including but not limited to movies, TV shows, audio and video programs, product catalogs, tourist guides, and video games. In one embodiment the system includes an "instant watch" mode and "wait and watch" mode, thereby allowing users to enjoy multimedia services instantly as the audio video program is being downloaded or after the download is substantially complete.