Deep Learning is thriving in recent years and finding increasing deployment thanks to the currently available processing power of computer systems. Big IT companies like Google or Facebook use Deep Learning algorithms in their daily business. Therefore, the reproducibility of research based upon Deep Learning algorithms is a crucial factor. This master thesis will focus on analyzing the influence of different operating systems as well as different Deep Learning frameworks. For this purpose, the same Deep Learning model is constructed and executed in three very popular frameworks (TensorFlow, Theano and Deeplearning4J). Further, different versions of these frameworks are considered as maybe some of them may implement crucial methods in a different way. Afterwards, the model is executed on seven operating system versions. Additionally, different versions of the used execution platform (Python and Java) will be considered. Finally, this thesis focuses on analyzing all obtained model results and testing if the results are significantly different when changing the execution context.