We investigate how children leverage intonational and gestural cues to an individual's belief state through unimodal (intonation-only or facial gestureonly) and multimodal (intonation + facial gesture) cues. A total of 187 preschoolers (ages 3-5) participated in a disbelief comprehension task and were assessed for Theory of Mind (ToM) ability using a false belief task. Significant predictors included Age, Condition and success on the ToM task. Performance improved with age, and was significantly better for the multimodal condition compared to both unimodal conditions, suggesting that even though unimodal cues were useful to children, the presence of reinforcing information for the multimodal condition was more effective for detecting disbelief. However, results also point to the development of intonational and gestural comprehension in tandem. Children that passed the ToM task significantly outperformed those that failed it for all conditions, showing that children who can attribute a false belief to another individual may more readily access these intonational and gestural cues.