Design-based research (DBR) is a diversified research genre: The combination of two worlds – that of research and that of education – and the different backgrounds and intentions of those involved entail different emphases, epistemological ideas, ideas on valuable outcomes and normative claims. This becomes visible in very different kinds of reasoning: No uniform structure of argumentation can be discerned, and so far, a differentiation into clear DBR types has not been convincingly achieved. This is a challenge for the orientation of DBR novices, the quality review of DBR studies, and the legitimation of DBR in the field of educational research. This article provides an empirical contribution to the discussion on argumentation: In a literature review, DBR studies are examined regarding their outcomes, the rationales authors use to justify their outcomes and indications for specific challenges in DBR reasoning. The analysis confirms for the sample that preliminary, prescriptive theory is most common alongside diverse practical outcomes. Authors often justify them with emphasis on variation, iteration, cooperation, and data triangulation. Different (standard) orientations, multi-level reasoning, and sub-studies present challenges for authors and readers, going back to the complexity of DBR projects. To justify their results in a comprehensible way, authors are confronted with the task to actively select an argumentation strategy.