The aim of this essay is to argue for the following claims: evil is prevalent; its prevalence is mainly the result of habitual and predictable patterns of action; these actions follow from the vices of their agents; in many cases, neither the evil actions nor the vices from which they follow are autonomous; it is nevertheless justified to hold the agents who perform these actions morally responsible for them; the widespread denial of this claim rests on the principle “ought implies (...) can”; two versions of this principle must be distinguished; neither version can be used to exempt agents from moral responsibility for their nonautonomous actions and vices; this has fundamental implications for how morality and responsibility should be conceived. (shrink)