On the face of it, it seems ludicrous to hold non-human animals morally responsible for what they do or don’t do. Ascribing moral agency to non-human animals seems like a category mistake – an example of what Fisher terms ‘categorical anthropomorphism’. In this talk, I attempt to unravel this belief by means of three case studies. A first case study is the historical practice of animal trials. In late medieval and early modern Europe, various kinds of animal wrongdoing were persecuted and punished. Canon laws were promulgated to excommunicate hungry caterpillars, wood mice, and slugs, while civil laws sentenced overly aggressive pets and livestock to death. A second case study is the late-nineteenth century scientific literature on animal criminality. Encouraged by Darwin’s claim that some animal behaviours and mental states, including emotions, are very similar to their human pendants, some criminologists started wondering whether perhaps non-human animals could also be charged with, among many other things, cannibalism, infanticide, and homosexuality. A third case study is a set of personal observations of humans, including myself, communicating with domesticated animals about their wrongdoings, in which I focus on moral emotions, such as anger, pride, and guilt. Tracking the similarities and differences between these three case studies, I chart the assumptions they make about the moral capacities of non-human animals, and I end with a compatibilist note on how to think about dealing with obnoxious animals.