An important characteristic of necessity cases is that ex ante and sometimes even ex post both parties have an interest in upholding the contract. From an economic perspective, the policy objective of regulating necessity is to give optimal incentives for precaution, search and rescue. These incentive effects have been widely discussed in the law and economics literature, the received view being that price control based on the costs of the rescue plus a small reward provide optimal incentives. In this paper we argue that the received view is unwarranted. Our model suggests that in many cases the socially efficient contract price is higher than the rescue costs. To be sure, due to serious information problems the practical implementation of this theoretical optimum is much more difficult than the cost-plus price setting supported by the received view. While the policy implications of our model are rather tentative we suggest that the judicial costs of estimating policy variables should be taken into account. A second objective of this paper is to argue for an economic interpretation of the term ‘necessity’ in contract law. Instead of various substantive criteria suggested in the philosophical literature we suggest defining the term by working backwards from the possible remedies. The excuse of necessity should be available for contracting parties when, all things considered, a judicial control (modification) of the contract price is desirable. Necessity is thus “defined” by what courts can and should do.