The major goal of this research project is to develop a ‘translatability prediction system’ for English-Dutch that not only automatically assigns a global difficulty score to a given source text, but also identifies the passages in the source text that are difficult to translate. The project addresses the following research questions: What are the typical source text segments that pose problems for translation? Do human translators on the one hand and machine translation systems on the other hand struggle with similar source text segments? How can we measure translation difficulty? What source text features can be used to predict translation difficulty?