This paper utilizes videos of naturally occurring interaction as its corpus and, by adopting a Systemic Functional multimodal discursive analytic perspective, proposes a tripartite model for studying the modal interaction in verbal gestural refusals. It thoroughly examines the modal interactional characteristics of children’s verbal gestural refusals from the three dimensions of physical time, cognitive mapping, and semiotic fusion. It reveals that children under examination demonstrate a preference for modal cooccurrence in temporal conversational sequences, prioritize transitivity mapping in iconic projection, and exhibit a balanced form of reinforcement and complementarity in modal integration. The strategy of cross modal interaction in physical, cognitive, and semiotic dimensions represents a typical pathway for children in conducting verbal gestural refusals, showcasing the interactive communicative competence of Mandarin children in verbal gestural interactions.