This paper presents a solution to the problem of designing effective systems for ontology-based data access. The solution is based on three main ingredients: a new ontology language based on Description Logics, a novel mapping language that addresses the impedance mismatch problem, and a query answering method that combines reasoning with specific mechanisms for accessing data. The ontology language, DL-Lite_A, is designed to handle large amounts of data and allows for efficient query answering by delegating the data-dependent part of the process to a relational DBMS. The mapping language enables the creation of new object identifiers from data values and addresses the impedance mismatch between data items and objects. The query answering method fully takes into account both the ontology and the mappings, and runs in LOGSPACE with respect to data complexity. The paper also discusses the semantics of DL-Lite_A ontologies and presents a framework for linking external data sources to an ontology expressed in DL-Lite_A. The results show that DL-Lite_A allows for efficient query answering and reasoning, and that the problem of linking data to the ontology can be addressed through the proposed mapping mechanism.This paper presents a solution to the problem of designing effective systems for ontology-based data access. The solution is based on three main ingredients: a new ontology language based on Description Logics, a novel mapping language that addresses the impedance mismatch problem, and a query answering method that combines reasoning with specific mechanisms for accessing data. The ontology language, DL-Lite_A, is designed to handle large amounts of data and allows for efficient query answering by delegating the data-dependent part of the process to a relational DBMS. The mapping language enables the creation of new object identifiers from data values and addresses the impedance mismatch between data items and objects. The query answering method fully takes into account both the ontology and the mappings, and runs in LOGSPACE with respect to data complexity. The paper also discusses the semantics of DL-Lite_A ontologies and presents a framework for linking external data sources to an ontology expressed in DL-Lite_A. The results show that DL-Lite_A allows for efficient query answering and reasoning, and that the problem of linking data to the ontology can be addressed through the proposed mapping mechanism.