The purpose of this article is to look at argument mapping in intelligence analysis and make suggestions for improvements in terms of analytic rigor and clarity, as well as justification when there is time to evaluate the boxes. Argument mapping is described in a similar way in the intelligence literature, but somewhat differently compared to philosophical literature and there are some things that are questionable or need to be clarified. It is also not clear what should be included in terms of analysis of competing hypotheses (ACH) or Bayesian analysis, for instance. The point of argument mapping is clarity of structure. Therefore, there should be a main claim or main hypothesis at the top, which is not an argument for something else in the tree, and which is argued for in the tree. ACH and Bayesian analysis should be performed before the argument map in order to find main hypotheses for separate trees. Even if it might be possible to put numbers on some boxes in the tree, doing it on all boxes might produce misleading results, depending on what they contain. The argument tree should be as clean as possible. Without numbers and likelihoods, we might use the notion of justified belief when investigating the tentative judgments so common in intelligence analysis.