Nowadays performance is very important in the software business. For example, if the search of an online shopping website takes too long, the customers won’t buy and the web site loses money. For measuring and optimizing performance, there are various solutions available. In this thesis, the focus is set on so-called profilers, more precisely on profiler data from YourKit, which is one of the leading tools in this segment. Profilers are used in development and can monitor all runtime data during the execution of a program. It measures for example the response time and saves the exact CPU and memory usage at any given time. The main aspect of this thesis is to analyze and detect different performance antipatterns in the profiler’s data export. Anti-patterns are the opposite of programming patterns, which are capturing expert knowledge of ”best practices“ in software design. Anti-patterns, on the other hand, document common mistakes made during software development. The goal is to automatically detect performance anti-patterns in the profiler data and show what the problem is and where it occurs. Therefore, this research is conducted with a company operating in the open-source domain. Together with them, we made a case study about the manual detection of anti-patterns in profiler data from load tests. This data is used in order to develop analysis strategies for the detection of anti-patterns with the help of a program called PADprof, which is also been developed in this thesis. The results show that most of the selected performance anti-patterns can be automatically detected in the available data. Nevertheless, more tests need to be conducted in order to evaluate if the anti-patterns can be detected in different data from other systems.