Although journalistic translation research has been quite successful over the past 15 years, from a methodological point of view many scholars struggle with the total or partial absence of a traceable source text. As a consequence, parallel corpora are rare and the researcher often has to rely on multilingual sets of texts that are comparable. This contribution deals in detail with that essential methodological problem. It relates the multi-source and multi-author situation of translation in journalism to this non- (or only partially) identifiable character of the source text–target text relationship. We argue that the triangulation of comparative text analysis with fieldwork adds value to this type of research. This argument is illustrated with a study triangulating textual analysis in three languages with interviews and non-participant observation. Such a triangulation also responds to earlier calls for a more elaborated contextualization of the production process and the socio-historical circumstances in journalistic translation research.