Two studies (a cross-sectional survey of 90 U.K. workers and an experiment with 100 U.K. workers) examined the cognitive and behavioral effects of abusive supervision. Both studies confirmed the hypothesis that workers who experience abusive supervision show paranoia and that makes them more prone to a type of cognitive error called the “sinister attribution error.” This is where workers misattribute innocent workplace events such as tripping over something or hearing colleagues laughing to malevolent motives such as wanting to harm or mock them. Study 1 also showed that abusive supervision is associated with lower well-being. Perceived organizational support buffers these effects, and this is associated with workers making fewer sinister attribution errors, thereby protecting well-being. Study 2 explored the role of contextual cues by exposing workers to images of abusive supervision. This increased their paranoia and contributed to workers making sinister attribution errors when they were asked to interpret workplace events. Moreover, depending on the types of contextual cues, workers were more likely to express the intention of workplace deviance after thinking about past experiences of abusive supervision. We recommend that corporate ethical responsibilities include training managers and workers about the negative cognitive and mental health effects of abusive supervision.