News translation has been investigated more systematically since the mid-2000s. Since then, it has kept pushing the boundaries of translation studies by asking such questions as the following: Can we study multilingual practices that do not necessarily have an identified source text? If so, what do we analyze and compare? Can we call these practices ‘translation’? How do we integrate multimodality into our traditionally textual analyses? This article formulates tentative answers based onrecent publications in the field. It calls for sustained research in the reception of news translation and with cognitive methods, as well as intensified exchanges with audiovisual translation.