The computer performs the hashing algorithm on the key field of the record, resulting in an address of a block and the computer stores the record in that block. If that block is already full, then the computer will store the record in an overflow area. A pointer will be left at the original block to indicate that a record is stored in the overflow area.

Eventually the blocks will become full and there will be a large number of records in the overflow areas. When this happens, the file will need to be re-organised to prevent access to records becoming too slow.

Access to records follows the same pattern...

• use the hashing algorithm on the key to find the address of the block where the record is stored.
• look for the record at the address. If it is not there follow the pointer to the overflow area.

Access to records is extremely fast. 

A random file would have to be stored on a direct access backing storage medium eg magnetic disc, CD, DVD

Example : Any information retrieval system. Eg Train timetable system.