Emotional information can be conveyed by various means of communication, such as propositional content, speech intonation, facial expression, and gestures. Prior studies have demonstrated that inputs from one modality can alter perception in another modality. To evaluate the impact of emotional intonation on ratings of emotional faces, a behavioral study first was carried out. Second, functional magnetic resonance (fMRI) was used to identify brain regions that mediate crossmodal effects of emotional prosody on judgments of facial expressions. In the behavioral study, subjects rated fearful and neutral facial expressions as being more fearful when accompanied by a fearful voice as compared to the same facial expressions without concomitant auditory stimulus, whereas no such influence on rating of faces was found for happy voices. In the fMRI experiment, this shift in rating of facial expressions in presence of a fearfully spoken sentence was correlated with the hemodynamic response in the left amygdala extending into the periamygdaloid cortex, which suggests that crossmodal effects on cognitive judgments of emotional information are mediated via these neuronal structures. Furthermore, significantly stronger activations were found in the mid-portion of the right fusiform gyrus during judgment of facial expressions in presence of fearful as compared to happy intonations, indicating that enhanced processing of faces within this region can be induced by the presence of threat-related information perceived via the auditory modality. Presumably, these increased extrastriate activations correspond to enhanced alertness, whereas responses within the left amygdala modulate cognitive evaluation of emotional facial expressions.