To search for specific elements in a marked up document we have, at least, two options: XPath and XQuery. However, the learning curve of these two dialects is high, requiring a considerable level of knowledge. In this context, the traditional Query-by-example methodology (for Relational Databases) can be an important contribute to make easier this learning process, freeing the user from knowing the specific query languages details or even the document structure. In this paper, we describe how we implement Query-by-example in a Web-application for information retrieval from a collection of structured documents, the GuessXQ system. In essence, we built an engine capable of deduce, from a specific example, the respective XQuery statement. After inferring the generic statement, the engine applies it to all documents in the collection to perform the desired retrieval. A suitable interface allows the end-user to mark over a sample document, picked up from the collection, the path he wants to select