This paper describes an environment for supporting very large ontologies. The system can be used on single PCs, workstations, a cluster of workstations, and high-end parallel supercomputers. The architecture of the system uses the secondary storage of a relational data base system, efficient memory management, and (optionally) parallelism. This allows us to answer complex queries in very large ontologies in a few seconds on a single processor machine and in fractions of a second on parallel super computers. The main contribution of our approach is the open architecture of the system on both the hardware and the software levels allowing us easily to translate existing ontologies for our system’s use, and to port the system to a wide range of platforms.