Recent discussions of contractualism and permissible risk imposition argue that because contractualism holds aggregate welfare to have no bearing on how it is permissible for individuals to relate to one another, it ends up committed to the position that faced with a choice between curing a person who has contracted a fatal disease and vaccinating a great many others, each of whom faces a significant risk of contracting it, curing is the only permissible option. But that is implausible. We should choose to the course that will result in fewer deaths.

I will argue that this charge gets things the wrong way around. When the choice between curing and vaccinating makes a difference to how many are expected to live, contractualist reasoning in fact supports the conclusion that it is impermissible to do other than vaccinate. What proves difficult to make sense of on its terms is the plausibility of curing rather than vaccinating in cases where the choice makes no difference to how many are expected to live. I propose a way of doing so that goes beyond contractualist thinking about permissibility, drawing on a more general ideal of how we ought to relate to one another animating the account.