In this paper we present a technique to access the contents of large XML documents. This technique can be applied to domains where large amounts of data are handled, on condition that those contents only require to be updated in a scheduled or sporadic way. We have implemented it in a system, named ALEP (Access to XML contents of Large Electronic Publications). First, ALEP preprocess the XML document, then it is used together with DOM API. One of their advantages is that it can be integrated into any software for management XML electronic publications. We have compared ALEP with DOM and SAX API’s improving both, time and space measures, with large documents of size greater than 18 MB megabytes acceding to a hundred of XML elements.