Adverse event reporting is a frequently used mechanism to establish a learning cycle to avoid future errors. However a precondition is that the potential - as well as the occurred adverse events is reported. This study explores two comparable internal medicine departments to find possible explanations on a differing frequency of adverse event reporting. Ethnographic methods - observation and interviews - are applied to collect data. The analysis reveals specific, but common ways of doing task prioritization and rating of adverse event severity. The interpesonal relationships, however, show significant differences in the two departments and can be the most plausible explanation of the difference in adverse event reporting.