Database systems currently storing legacy data can be enhanced to leverage all of the benefits that a global character set has to offer by migration to Unicode. This can be achieved but often not without limitations and unforeseen problems. Careful planning and consideration are invariably overlooked increasing the long term maintenance costs.

This paper outlines some new techniques for migrating a database to Unicode where focus is placed on data analysis and issue identification. Solutions to related issues, such as data loss and integrity, are also considered in ensuring this often complex operation is executed without problems.

At the conclusion, an example of a simple migration will be demonstrated using existing Oracle technology.