We investigate the effects of babyfaceness on the trustworthiness and judgments of a company's chief executive officer in a public relations crisis. Experiment 1 demonstrates boundary conditions for the babyfaceness‐honesty trait inference and its influence on company evaluations. Experiment 2 shows that trait inferences of honesty are drawn spontaneously but are corrected in the presence of situational evidence (a severe crisis) if cognitive resources are available. We demonstrate that these babyface‐trait associations underlie evaluations by reversing the babyface effect on judgments in (a) experiment 3, where a priming task creates associations counter to the typical babyface–unintentional harm stereotype, and (b) experiment 4, which creates a situation where innocence is a liability.