The software testing discipline has gained increasing importance over the past 30 years and is nowadays a vital part in every software project. To enhance efficiency and simultaneously reduce costs effort is put in the automation of recurring manual software testing tasks. The graphical user interface of a system plays a crucial role in establishing interaction between the system-s underlying functions and the user. Despite this fact, testing the graphical user interface is a complex and time-consuming task and automation strategies are still in an initial stage. Only rare information on the availability and maturity level of test automation tools for Java based rich client systems and on their suitability for practical application is available. This thesis overcomes this deficiency in knowledge by investigating six actively developed test automation tools, which are licensed as open source and are suitable for testing the graphical user interface of Java based desktop applications. To perform this investigation, the application of the selected tools within a representative example project is required. Due to its scope, level of complexity as well as usage of state of the art technologies the rich client of an existing event management system is chosen for this task. Object of study is the functional scope of the tools as well as non-functional quality parameters such as integrability, robustness and maintainability of the test scripts. Additionally, their suitability for practical application on the existing software project is analyzed. First, background information on the concept of graphical user interfaces as well as the discipline of software testing is provided. In order to select the best candidates, a criteria catalog containing general and project specific requirements for test automation tools is developed and evaluated. Following this, six automation tools, namely FEST, Marathon, Abbot, Sikuli, Robot Framework and WindowTester Pro, are chosen for further investigation. In the main part of this thesis a feasibility study is conducted in which the functioning of each test automation tool is analyzed by implementing five representative test cases from the existing software project. Furthermore, advantages as well as weaknesses of each tool are discussed and possible improvements are introduced. It is concluded, that none of the tools investigated is ready for practical application without extensive adaption work prior to usage. The results obtained from this feasibility study are due to the representative character of the chosen example project also highly relevant for inexperienced testers or small companies that cannot afford licenses for commercial test automation tools. They can serve as a decision guideline when introducing test automation in similar projects.