A system and method for identifying an item in a video program including associated metadata. In operation, the correspondence between product information identifying the item, and a scene depicted in the video program in which the product appears, is stored in a database. The associated metadata, previously embedded in the video program, is also copied to the database, and indicates a correspondence between each said scene depicted in the program and the location therein of the scene. A specific scene of interest in the video program is identified by reference to the metadata in the database, and the item in the specific scene is then identified by correlating the metadata for the specific scene with the product information stored in the database, using the stored product/scene correspondence.