Microsomal prostaglandin E synthase-1 (mPGES-1) is an essential enzyme involved in a variety of diseases and is the most promising target for the design of next-generation anti-inflammatory drugs. In order to establish a solid structural base, we recently developed a model of mPGES-1 trimer structure by using available crystal structures of both microsomal glutathione transferase-1 (MGST1) and ba3-cytochrome c oxidase as templates. The mPGES-1 trimer model has been used in the present study to examine the detailed binding of mPGES-1 trimer with substrate PGH(2) and cofactor GSH. Results obtained from the computational alanine scanning reveal the contribution of each residue at the protein-ligand interaction interface to the binding affinity, and the computational predictions are supported by the data obtained from the corresponding wet experimental tests. We have also compared our mPGES-1 trimer model with other available 3D models, including an alternative homology model and a low-resolution crystal structure, and found that our mPGES-1 trimer model based on the crystal structures of both MGST1 and ba3-cytochrome c oxidase is more reasonable than the other homology model of mPGES-1 trimer constructed by simply using a low-resolution crystal structure of MGST1 trimer alone as a template. The available low-resolution crystal structure of mPGES-1 trimer represents a closed conformation of the enzyme and thus is not suitable for studying mPGES-1 binding with ligands. Our mPGES-1 trimer model represents a reasonable open conformation of the enzyme and is therefore promising for studying mPGES-1 binding with ligands in future structure-based drug design targeting mPGES-1.