Most people believe that the rights of others sometimes require us to act in ways that have even substantially sub-optimal outcomes, as viewed from an axiological perspective that ranks outcomes objectively. Bringing about the optimal outcome, contrary to such a requirement, is an ‘optimific wronging’. It is less clear, however, that we are required to prevent optimific wrongings. Perhaps the value of the outcome, combined with the relative weakness of prohibitions on allowing harm as compared to those against doing harm, justifies non-intervention. In this article, I consider arguments to that effect, focusing on a recent paper in this journal by Andreas Mogensen. I argue that while we do not, in general, do wrong by failing to prevent optimific wrongings, we are nevertheless not permitted, in key cases, to refrain from intervening on the grounds that not intervening will secure the optimal outcome.