In this paper we introduce the OMAR (Object Modeling of Relational Applications) methodology, which uses object oriented front end modeling and supports the generation of relational applications through the automatic inference of the relational database schema and the structure of application programs from the outputs of the object oriented design phase. OMAR also includes an object/relational programming environment, the PC++, an extension of C++ with abstraction mechanisms to simplify the connectivity with a relational DBMS within the object oriented programming paradigm. In  methods for mapping object schemes onto relational schemes have been given. In that, mapping rules were used to translate class diagrams into relational data schemes. As opposed to this method, OMAR uses not only the information captured within the class diagrams, but also the information contained within the dynamic models (State Diagrams and Data Flow Diagrams), which is used by a Prolog based inference engine to derive a more appropriate relational data schema and to automatically generate the manipulation code. This also provides considerable advantages as opposed to traditional waterfall system development methodologies, where the data and functional modeling were two distinct phases and were often conducted by different groups, which entailed a tedious data/function integration phase. We relied on ODBC to provide compatibility with more relational platforms. The paper is organized as follows. In Section 2 we discuss some background issues and introduce the OMAR development process. The mapping technique from the object model to the relational model together with a case study is described in Section 3. In Section 4 we describe the OMAR programming paradigm using the PC++ programming language and provide the rules for mapping object schemes onto PC++ programs. Finally, conclusions are given in Section 5.