Objective: To investigate the impact of pediatric traumatic brain injury (TBI) on multisensory integration in relation to general neurocognitive functioning. Method: Children with a hospital admission for TBI aged between 6 and 13 years (n = 94) were compared with children with trauma control (TC) injuries (n = 39), while differentiating between mild TBI without risk factors for complicated TBI (mildRF−; n = 19), mild TBI with ≥1 risk factor (mildRF+; n = 45), and moderate/severe TBI (n = 30). We measured set-shifting performance based on visual information (visual shift condition) and set-shifting performance based on audiovisual information, requiring multisensory integration (audiovisual shift condition). Effects of TBI on set-shifting performance were traced back to task strategy (i.e., boundary separation), processing efficiency (i.e., drift rate), or extradecisional processes (i.e., nondecision time) using diffusion model analysis. General neurocognitive functioning was measured using estimated full-scale IQ (FSIQ). Results: The TBI group showed selectively reduced performance in the audiovisual shift condition (p = .009, Cohen’s d = −0.51). Follow-up analyses in the audiovisual shift condition revealed reduced performance in the mildRF+ TBI group and moderate/severe TBI group (ps ≤ .025, ds ≤ −0.61). These effects were traced back to lower drift rate (ps ≤ .048, ds ≤ −0.44), reflecting reduced multisensory integration efficiency. Notably, accuracy and drift rate in the audiovisual shift condition partially mediated the relation between TBI and FSIQ. Conclusion: Children with mildRF+ or moderate/severe TBI are at risk for reduced multisensory integration efficiency, possibly contributing to decreased general neurocognitive functioning.