A protein model (such as a ribbon model) can be created from the atomic coordinates in the protein data base files. These coordinates are obtained by X-ray crystallography or NMR spectroscopy with the protein arbitrarily oriented. As such, identifying or comparing a novel structure with a known item using protein model in the protein data base can be a timely process since a large number of transformations may be involved. The identification efficiency will be improved if the protein models are uniformly oriented. This paper presents an approach to orient a protein model to generate the characteristic views with minimum and maximum crossings respectively. The projection directions for these characteristic views are determined by a set of crossing maps (C-maps). Re-orientating the protein models in the protein data base to two characteristic views will facilitate the process of identification.