A method for realizing multi-picture, including: A. Multipoint Control Unit (MCU) creates an indicator containing multi-picture information for each of the terminals and sends the indicators to the terminals; B. each terminal receives the indicator, encodes a source picture depending on the information in the multi-picture information indicator, and inserts said multi-picture information indicator into a sub-bitstream and transmits the sub-bitstream to the MCU; C. when receiving the sub-bitstreams from the respective terminals, the MCU combines the sub-bitstreams into a composite bitstream and sends the composite bitstream to a display terminal; D. the display terminal receives the composite bitstream, extracts the sub-bitstreams from the composite bitstream, and decodes the sub-bitstreams depending on the multi-picture information indicators of the sub-bitstreams and combines the decoded sub-pictures into a multi-picture output. The present invention reduces the cost and greatly enhances multi-picture transmission and combination capability thus to implement transmission and combination of sub-pictures.