OBJECTIVES: Adverse events –Aes - due to medication errors are numerous. One type of error is the combination of substances with a known unwanted pharmacological interaction. In the years 2001 to 2006 a total of 39521 AEs with ICD10 codes indicating drugs as reason where documented in Austrias hospitals.  There is no reporting system or documentation for adverse events in ambulatory care. So the rate of AEs in the outpatient setting is unknown. One of the main projects in Austrias e-health roadmap is an e-medication system. Due to the foreseeable costs an estimation and calculation of potential benefits is mandatory. METHODS: We use reimbursement data of the Austrian social insurance companies covering the years 2006 and 2007. The domains included are pharmaceutical reimbursement, reimbursement of outpatient medical services, hospital reimbursement data and sick leave data. We focus on interactions between the following pharmaceutical groups: 1. ATC code C10AA – statins, 2. ATC code N5BA – benzodiacepins, 3. ATC code J01FA makrolide antibiotics and 4. ATC code C01BD amiodarone. RESULTS: The main question to answer is the prevalence and incidence of such interaction patterns. This will lead to a description of the epidemiology of potential AEs which can at least be qualified as medical errors. The second question is the frequency of clinical relevant interactions. A clinical (and economic) relevance is ascertained, if the interaction needs medical treatment. The detection of outcomes of an interaction is the main challenge. CONCLUSIONS: The extent to which an e-medication system can improve patient safety can be estimated. Therefore prescriptions causing interaction coming from one office should be prevented. But prescriptions from different offices cannot be avoided by the use of this isolated solution. Only a comprehensive e-health solution summoning up medication of a patient from different sources has the potential to improve patient safety. There are pharmacologic interactions with different clinical relevance.