Sometimes, an action has something morally in its favour, but it is not morally required since an alternative action is favoured by moral considerations that have, in the context, a greater normative weight. Cases of this sort suggest that it is one thing for an action to be favoured by a moral consideration, or reason, and quite another for it to be morally required, or obligatory. In view of this, it makes sense to ask how moral reasons and moral obligations are related to one another. In my talk, I suggest and defend an answer to this question.
More specifically, I show that there is an initially plausible and theoretically defensible way of explaining moral obligations in terms of moral (and other) reasons, i.e. of complementing the following scheme:

There is a moral obligation to φ if, and only if (and because), … [moral (and other) reasons].

As has been observed by several authors, the key to accomplishing this task is to make appropriate room for the possibility of supererogation. I argue that the view that I propose does this in a particularly promising way: it turns out to be not only consistent with the possibility of supererogation, but to entail a plausible general account of what makes an action supererogatory in the first place. Moreover, it can be shown to entail an attractive view about how deontic moral categories and all-things-considered categories are related to one another.