We had participated in the CASP (Community Wide Experiment on the Critical Assessment of Techniques for Protein Structure Prediction) that is the competition of the protein structure prediction in 2000, 2002 and 2004. Our server were estimated almost the best as the individual homology modeling server in the CAFASP3 (Critical Assessment of Fully Automated Structure Prediction) that is the full automatic homology modeling section of CASP5. In this competition we used homology-modeling software FAMS[1] (Full automatic protein modeling system) developed for several years[2]. CAFASP4 results will come soon in last of 2004. Thus, We hope to show the usefulness of FAMS clearly in this competition. We also have developed FAMSBASE that is FAMS model structure database[3]. In this article we describe that proteins coded in the 277 species genomes were modeled in FAMSBASE.