Indexing is often explained as a two step procedure whereby indexers first analyze the document to determine its subject matter and then translate the subject matter into index terms. In the domain-centered approach, these two steps are the last two steps in a series of analyses that begins with an analysis of the domain and includes an analysis of the users information needs as well as the indexers’ roles. Although the two step procedure is beneficial in separating the analysis of the document’s subject matter and the creation and selection of index terms, it is insufficient in laying out the steps that an indexer must go through when indexing a document. The analysis of documents to establish their subject matter is a complex phenomenon and it requires complex explanations to guide the indexers. The domain-centered approach to indexing offers a framework that manages the complexity and guides the indexer through the range of required analyses.