Software today is becoming increasingly more complex and extensive, and its documentation is becoming more and more complicated. During the software life cycle documentation tends to accumulate a lot of duplicates due to copy-pasting: first, some text fragment is copied, possibly multiple times, then each copy is modified, possibly in its own way; thus, different copies of the initial fragment become near duplicates. Depending on the document type, duplicates can be either desired or not. In either case, they need to be managed during documentation lifecycle. This paper introduces a duplicate management process and shows how it could be applied in documentation maintenance. An example of duplicate management for industrial documentation is presented.