Argument maps represent some arguments more effectively than others. The goal of this article is to account for that variability, so that those of who wish to use argument maps can do so with more foresight. I begin by identifying four properties of argument maps that make them useful tools for evaluating arguments. Then I discuss four types of argument that are difficult to map well: reductio ad absurdum arguments, charges of equivocation, logical analogies, and mathematical arguments. The difficulties presented by these four types appear unrelated to one another, but I show that, in each case, the difficulty can be traced back to the use of metalinguistic reasoning. The need to represent a transition between object language and metalanguage can undermine one or more of the benefits that argument map representation would otherwise confer.