In my paper, I address the – actual and potential – structure of comparative law arguments in legal reasoning. By ‘comparative law arguments’ I mean arguments that feature as one of their premises a comparative relation, that is, a relation of similarity or difference, between the laws of non-identical legal systems. And by ‘legal reasoning’ I capture reasoning about what the law is as well as reasoning about how a case ought to be decided according to the law (as opposed to how a case ought to be decided all things considered). I start by analysing actual, and prominently discussed, instances of comparative law arguments. The three instances I pick out are (1) the argument from consensus among legal systems worldwide to prohibit the juvenile death penalty as featured in the US Supreme Court’s decision in Roper v. Simmons, (2) the so-called floodgate argument and its refutation encountered in national and international debates about the protection of pure economic loss in tort law, and (3) an argument that draws on the fact that another legal system solves a similar legal issue in a certain way to conclude that this solution would solve the similar legal issue in this legal system just as well. I demonstrate that (1) is best rendered as an (inductive) argument from authority, that (2) may be construed as a deductive inference, and that (3) is an instance of analogical reasoning. Apart from varying modes of inference, comparative law arguments feature different types of conclusions. In a second step, I examine under which circumstances and presuppositions we may reliably infer useful insights for legal reasoning from comparative relations between the laws of non-identical legal systems. While such comparative relations will only rarely allow for reliable conclusions about what the law is, they may frequently (be construed to) amount to strong reasons to believe that a certain decision is the best one according to the law.