Representing genomic mapping knowledge is a complex process due to the diversity of experimental approaches. The lack of obvious equivalence between different kinds of maps¹ makes it difficult to use them simultaneously or in a cooperative way. Furthermore, comparative mapping is not limited to map comparison, it also implies the comparison of the elements they are composed of. Comparison thus implies the combination of different computational objects, corresponding to different levels of information. However, from the user’s point of view, systems devoted to comparative mapping should appear to process queries in a homogeneous manner, whatever the objects they imply. We propose a data model that integrates mapping data, genetical data and sequences and takes into account the specific constraints resulting from the comparison of maps between and within different species.