3D complex structures of proteins and other molecules provide a clue for mechanism of interaction. However, few servers for searching complex structures are available. We are developing an updated server HOMCOS (HOmology Modeling of Complex Structure) with more services than the previous version. It separates component molecules of PDB files of complexes, such as proteins, nucleic acids, small chemical compounds, stores their binding relationships. It searches these molecules by BLAST and our chemical structure comparison program dkcombu. Based on found similar complexes, simple template-based models of the complex can be generated by replacing the template with the query amino acid sequence. The service ‘Modeling Homo Protein Multimer’ searches homologous protein complexes with one query protein sequence, whereas the service ‘Modeling Hetero Protein Multimer’ searches with two sequences. The service ‘Modeling Compound-Protein Complexes’ searches with one sequence and one chemical structure. For modeling a bound conformation of a target compound, our flexible overlay program fkcombu is performed to superimpose the target compound onto the template compound bound to a protein. The program transforms a pose and torsion angles of the target molecule to superimpose atom pairs of the target and template molecule determined by 2D-MCS. The service ‘Searching Contacting Molecules for Query Protein’ searches contacting molecules with homologous proteins to a given query sequence, and provide putative interacting molecules with putative binding sites. It is useful to estimate mutational effects on molecular interactions. The service ‘Searching Contacting Proteins for Query Compound’ searching contacting proteins with similar compounds to a given query chemical compound. It may be useful to predict unintended interacting proteins, which cause side effects of the compound. We hope this server may contribute to the platform for drug discoveries and life sciences.