Sometimes grave harm is an aggregate effect of many actions' consequences. In such cases, each individual contribution may in fact be so tiny that it's effect are imperceptible and, therefore, can't be said to cause harm. This creates a problem: If no agent's contribution causes harm, then s/he may lack sufficient reason not to contribute to the overall harmful effect. Hence, the harmful effect would seem normatively inevitable. In this talk, I discuss and reject two broadly consequentialist answers to the problem and I defend a virtue ethicist solution. According to the former views, a thorough analysis of all relevant cases reveals that there really are no cases of imperceptible contributions to harm. According to (my favored version of) the latter approach, agents should (in part) not partcipate in harmful practices, because a reasonably emphathetic agent would not act in this way.