Clinical trials only allow for a limited assessment of adverse drug reactions (ADRs) due to their restricted number of patients, their short duration, and their narrow inclusion criteria. To assess potential ADRs in the population actually treated after drug approval, different systems for signal generation such as spontaneous reporting systems have been established and are required by drug laws in many countries. As of today epidemiological methods for signal confirmation for rare adverse events or events with long induction periods use record linkage databases such as the United Health Care (Ingenix) in the USA or the practice databases such as the General Practice Research Database in Great Britain. However, this approach also has its limitations. This type of database will be presented in this paper. In addition, databases operated by German public health insurance will be described. The topic of linking different data sources in Germany to support pharmaco-vigilance activities will be addressed.