Natural Language Interfaces to Databases (NLIDB) are used to facilitate non-technical users. Currently, data warehouses are widely used by enterprises for decision-making. The users of such decision-making systems are top management who are normally non-technical having less knowledge of the data warehouse schema and about writing technical queries. This paper proposes a Natural Language Retrieval System for Data Warehouse (NLRSDW) that facilitates executives to interact with Data Warehouse in Natural Language. This system uses the Online Analytical Processing (OLAP) engine to efficiently and accurately generate the results. The use of ME/R model, data graph, utilization of Materialized Views (MVs) and query optimization technique supports in good performance of the system. A comparative experimental analysis with existing systems shows that NLRSDW has reasonable stance.