Since the SPOUT superfamily was defined by homology between the SpoU and TrmD families [Anantharaman, V. et al., J. Mol. Microbiol. Biotechnol., 4, 71-75 (2002)], many crystal structures have been solved and numerous new homologous sequences have been found in the superfamily. Therefore, nowadays, we can consider enzyme function and/or evolution process of the SPOUT superfamily members using not only amino acid sequences but also protein structures. Recently, a bioinformatics research on SPOUT superfamily proposed existences of new member proteins [COG1756, COG1772, COG4080, and COG1901], and provided a structural and evolutionary classification of the proteins [Tkaczuk, K.L. et al., BMC Bioinformatics, 8:73 (2007)], which serves as a guide for studies on the SPOUT family in future. In this meeting, we report a new approach using a flexible protein structure alignment algorithm (FATCAT) to analyze the structures of SPOUT superfamily proteins, and discuss differences in substrate selectivities of methyltransferases in the superfamily.