Several matching procedures have been proposed over the years to improve the search time of a query sequence in a database of molecular sequences. However due to the architectural limitations of conventional computers, only one sequence can be accessed/retrieved at a time from a database. The use of a parallel processing at least in theory, can provide a tremendous boost to the comparison of sequences. However, no algorithm employing parallel operations for this purpose is currently available. This paper presents a content addressable memory-based approach for fast comparison of molecular sequences.