Abstract: Under the context of the “dual-carbon” targets, methane-ammonia blended combustion offers significant decarbonization potential for gas and power industries. Wall confinement in the combustor is a key factor influencing combustion characteristics; however, its impact mechanism on ammonia-blended combustion has not yet been fully elucidated. To clarify the effects of wall confinement on methane-ammonia swirl combustion, a combined approach of experimental measurements, computational fluid dynamics (CFD) numerical simulations, and chemical reactor network (CRN) analysis was employed to systematically investigate the influences of different wall confinement conditions on flame structure, flow characteristics, and NO formation behavior. The results show that as the confinement ratio increases from 5.8 to 14.9, the central recirculation zone expands significantly and the recirculation intensity is enhanced, accompanied by a reduction in main jet velocity and turbulent kinetic energy. Meanwhile, the flame gradually spreads outward, the flame area increases, and the local chemiluminescence intensity decreases markedly. In addition, the effect of wall confinement on NO emissions strongly depends on the combustion atmosphere. Under lean and near-stoichiometric conditions, increasing the confinement ratio can reduce NO emissions by more than 40%, whereas under rich conditions, the sensitivity of NO emissions to wall confinement is substantially weakened. CFD-CRN analysis further indicates that changing the confinement ratio does not fundamentally alter the relative contributions of dominant NO reaction pathways, but simultaneously weakens both NO formation and consumption rates due to the reduction in turbulence intensity and associated turbulent–chemical inter-actions. Furthermore, this study validates the reliability of the coupled CFD-CRN approach in predicting species concentrations, demonstrating its capability to accurately capture both the trend and magnitude of NO emissions, and providing an efficient, low-cost simulation tool for engineering applications.