Objective: This study aims to investigate the efficacy, safety, and brain network mechanisms of combined central–peripheral intermittent theta-burst stimulation (iTBS) for poststroke dysphagia (PSD). Methods: We randomly assigned 89 PSD patients in a 1:1:1 ratio to three groups: the swallowing cortex iTBS (SC-iTBS), suprahyoid muscle iTBS (SM-iTBS), and combined iTBS (C-iTBS) groups. Each group received the corresponding iTBS intervention. Swallowing function was assessed using the standardized swallowing assessment (SSA), penetration–aspiration scale (PAS), Yale pharyngeal residue severity rating scale (YPR-SRS), motor-evoked potential (MEP), and functional near-infrared spectroscopy (fNIRS) at baseline (T0) and after treatment (T1). Results: At T1, the SSA, PAS, YPR-SRS, and MEP amplitude scores in all three groups were significantly better than those at T0. The C-iTBS group improved significantly more than the single-intervention groups (p < 0.05). Significant differences were found in the efficacy rates among the three groups (SC-iTBS: 54.84%, SM-iTBS: 80%, C-iTBS: 85.71%, χ2 = 8.236, p = 0.016). The C-iTBS group presented a larger effect size (D: 1.249–2.149). The fNIRS results revealed significant differences in the mean cortical brain function connectivity strength, partial brain network connectivity strength, and brain region activation levels at T1 compared with those at T0. Compared with the other two groups, the C-iTBS group presented significantly greater functional connectivity strength and brain region activation levels in specific heteromodal brain networks (p < 0.05). No serious adverse events were reported during the entire trial. Conclusion: Dual-target iTBS (C-iTBS) was better than single-target interventions at improving swallowing function, potentially through enhanced functional connectivity and activation in heteromodal swallowing-related networks.