Abstract. Data privacy is an important application of ontology modularization. The aim is to publish one module while keeping the information of another module private. We show how locality and partitioning two basic concepts in the theory of modular ontologies - naturally lead to privacy preserving query answering over modular ontologies.