Abstract <p>Gas chromatography with mass spectrometric detection is widely used for non-targeted analysis of environmental samples, including the monitoring of ecotoxicants and their transformation products. For identification of detected components, not only libraries of experimental mass spectra are used, but also data obtained in silico. Authors developing new prediction approaches and comparing them with existing solutions, often focus on prediction accuracy. At the same time, the reproducibility of the approach and its usability for a wide audience are usually not discussed. As shown in the present study, the use of algorithms described in the literature may be limited because of the absence of predictive model weights or the project source code. Even when the code is available, the use of the algorithm may be problematic because of the outdated versions of the programming language and the libraries used. In the present study, the available approaches to prediction of electron ionization mass spectra are compared, with the main focus on ease of installation, execution and use, rather than on prediction accuracy. The problems are described in detail and solutions are proposed.</p>