A crossmodal effect has been observed in the processing of facial and vocal emotion in adults and infants. For the first time, we assessed whether this effect is present in childhood by administering a crossmodal task similar to those used in seminal studies featuring emotional faces (i.e., a continuum of emotional expressions running from happiness to sadness: 90% happy, 60% happy, 30% happy, neutral, 30% sad, 60% sad, 90% sad) and emotional prosody (i.e., sad vs. happy). Participants were 5-, 7-, and 9-year-old children and a control group of adult students. The children had a different pattern of results from the adults, with only the 9-year-olds exhibiting the crossmodal effect whatever the emotional condition. These results advance our understanding of emotional prosody processing and the efficiency of crossmodal integration in children and are discussed in terms of a developmental trajectory and factors that may modulate the efficiency of this effect in children.