In criminal cases, jurors are expected to evaluate evidence rationally, make objective causal inferences, and reach fair decisions that can profoundly affect the lives of both victims and defendants. However, psychological research suggests that this is not necessarily the case: when judging whether someone has caused a harmful outcome, individuals appear to be influenced, under certain conditions, by the moral valence assigned to the agent's motivations (i.e., whether these motivations are praiseworthy or blameworthy). The present research examines whether the moral valence effect on causal judgments is more likely to emerge under causal ambiguity, that is, when the agent's action is a necessary condition (but for the event) but not the most direct cause of the harm due to the presence of an intervening factor with a stronger causal link. In two pre-registered studies (N = 425), participants read short scenarios in which the moral valence of the agent's motivations (Studies 1 and 2) and that of the victim (Study 2) were orthogonally manipulated. They were then asked to judge the extent to which the agent was the cause of the negative outcome. Although the agent's action was not the actual cause of the outcome, the blameworthy (vs. praiseworthy) agent was more selected as the cause of the harm. However, a negative (vs. positive) evaluation of the victim reduced causal attributions to the agent, regardless of the agent's moral valence. Moreover, exploratory analyses may have revealed a "black sheep effect": men were less likely than women to attribute causality to a blameworthy agent. These findings deepen our understanding of how agent and victim evaluations shape causal judgments. They also reflect critical issues in legal contexts, where moral evaluations may implicitly influence causal attributions and blame, even in the absence of clear causal links.