The aim of this thesis is to provide the information that is needed to build a correct data history in a Business Intelligence System. Thereby the phenomena of "Slowly Changing Dimensions" is emphasised. At first a definition of business intelligence is provided which includes the ETL-process, data warehousing, and on-line analytic processing. Furthermore data mining, reporting, dashboards an BI-portals can be considered as a part of the BI-process. It is possible to represent slowly changing dimensions in a database management system. This is modelled in the ETL-process. The data is transformed in a certain way, before being loaded into the data warehouse. Type 1 SCD represents a simple update of the current dimension entry. Type 2 SCD provides a history of the data. It saves all the intermediate states of it. Thus, a user can track all states that changed during a certain time. Type 2 SCD is the one which is mostly used. Type 3 SCD provides just the original and the actual state. Aside from this three types there are existing some hybrid and extended versions.