Matching records in the destination list are replaced with corresponding source file records. New records can be added to the end of the list. The merge is based on an index, which is used to check for duplicate records which exist in both files.