The development and maintenance of Web sites are difficult tasks. To maintain the consistency of ever-larger, complex Web sites, Web administrators need effective mechanisms that assist them in fixing every possible inconsistency. In this paper, we present a novel methodology for semi-automatically repairing faulty Web sites which can be integrated on top of an existing rewriting-based verification technique developed in a previous work. Starting from a categorization of the kinds of errors that can be found during the Web verification activities, we formulate a stepwise transformation procedure that achieves correctness and completeness of the Web site w.r.t. its formal specification while respecting the structure of the document (e.g. the schema of an XML document). Finally, we shortly describe a prototype implementation of the repairing tool which we used for an experimental evaluation of our method.