In developing its chemical information system, Chemical Abstracts Service (CAS) has built files which require hundreds of millions of bytes of storage and in which data elements may range in length from zero in one instance to several hundred bytes in another. To cope with this size and variability, CAS has developed an internal standardized storage format and has implemented it on the IBM System/360. This format, which is called Standard File Format (SFF), accommodates a mix of fixed-and variable-length data elements and permits the addition and deletion of data elements without affecting the program code. Symbolic addressing of data elements is achieved through the use of a directory within each record in the file. Implementation of standardized files has permitted the CAS system to operate with greater generality and flexibility.