Schema matching attempts to discover semantic mappings between elements of two schemas. Elements are cross compared using various heuristics (e.g., name, data-type, and structure similarity). Seen from a broader perspective, the schema matching problem is a combinatorial problem with an exponential complexity. This makes the naive matching algorithms for large schemas prohibitively inefficient. In this paper we propose a clustering based technique for improving the efficiency of large scale schema matching. The technique inserts clustering as an intermediate step into existing schema matching algorithms. Clustering partitions schemas and reduces the overall matching load, and creates a possibility to trade between the efficiency and effectiveness. The technique can be used in addition to other optimization techniques. In the paper we describe the technique, validate the performance of one implementation of the technique, and open directions for future research.