Laws of nature have long been thought to have special significance for aspects of scientific reasoning such as counterfactual conditionals, inductive projections, and scientific explanations. But the laws' distinctive roles in scientific reasoning have proved notoriously difficult to identify precisely, leading some philosophers even to suggest that there are no such roles. The aim of this book is to determine these roles and see what a law of nature must be in order for the laws to function as they do in scientific practice. Lange shows that the laws possess a uniquely broad range of invariance under counterfactual perturbations, a range that for the first time is characterised without appealing to the concept of a law. It is argued that the laws fail to supervene on the nonnomic facts, just as the rules governing chess fail to supervene on the moves made in a given actual game. It is also argued, against both regularity accounts and analyses of laws as relations among universals, that a law need not be associated with an exceptionless regularity. It is explained how a law of one scientific field (e.g. cardiology) can be an accident of another (e.g. fundamental physics). Special attention is paid to laws of biology and other 'special sciences', and it is argued that their distinctive range of invariance allows these fields to supply scientific explanations that are irreducible, even in principle, to explanations in terms of fundamental physics. Another special feature of this book is its emphasis on the distinction between laws of nature and physically necessary coincidences, a distinction crucial to the concept of natural kind. An account is also given of 'meta-laws', such as symmetry principles. Among the philosphers receiving special discussion are Lewis, Goodman, van Fraassen, Armstrong, Dretske, Earman, Mill, Fodor, Hempel, Giere, Putnam, Dennett, and Mackie.