The Spatial Information Infrastructure (SII) Law has been in effect since 7th June, 2010. This law is an implementation of the INSPIRE Directive for Polish conditions. According to the 30th article of the law, the existing spatial data sets should begin adapting to the requirements of the law and new sets should be created accordingly. A lot of articles concerning the INSPIRE Directive, spatial data sets harmonization, integration of the sets with the ISO 19100 series of International Standards, sets interoperability, metadata etc. have been published in the specialist press. However, some significant elements connected with the INSPIRE Directive and its implementation for Polish conditions are missing. Those elements are data specifications and the INSPIRE implementation rules. The basic elements that should be taken into consideration during georeference database modeling to ensure the compatibility of the created models with the INSPIRE Directive guidelines and the SII Law, are specified and described in this paper. In the author’s opinion, the basics of database modeling are as follows: 1. Conceptual modeling as the fundamental modeling methodology; 2. UML – a formal language to write a model; 3. Taking into account the INSPIRE implementation rules when preparing models; 4. Taking into consideration the INSPIRE data specifications in created models; 5. Harmonization of prepared models in a given domain; 6. Integration of created models with the ISO 19100 series of International Standards. All of the above mentioned “parts” of the modeling process aim at ensuring the interoperability of databases, built on the basis of the prepared models, not only at the local level (the Polish administrative units, e.g. communal) but also at the national and the international levels. Ensuring interoperability of the spatial data sets that come from different countries is one of the main aims of the INSPIRE Directive. The characteristic of particular modeling stages is based, if possible, on concrete examples from photogrammetry and the remote sensing domain.