Excellent oxidation resistance is essential for the viability of high-entropy carbides (HECs) in high-temperature environments. However, the vast HEC compositional space presents a significant challenge in developing desirable formulations with improved oxidation tolerance. In this work, we establish a computational framework to guide the optimization of oxidation-tolerant HEC compositions and further validate the theoretical predictions through systematic oxidation experiments. Leveraging first-principles calculations, we initially determine the heterogeneous oxygen adsorption ability of different constituent elements in HECs, which is subsequently harnessed to regulate preferential oxidation and suppress the formation of detrimental oxides during the oxidation process. Incorporating this fundamental-level understanding, multi-objective optimization (MOO) was performed to identify a compositional region with potentially intrinsic oxidation immunity, achieving a balance between the compactness of the oxidation product and the thermodynamic stability of the HEC matrix. Based on our analysis, we propose a compositional design strategy involving the strategic reduction of elements with higher oxygen adsorption energies, such as Nb, Ta, and Ti in (NbTaZrHfTi)C, to effectively improve the oxidation performance of HECs. The theoretical findings are robustly corroborated by our comparative oxidation experiments, which demonstrate that the selected NbTa-depletion HEC exhibits superior antioxidant performance compared to the equiatomic counterpart. The integrated computational-experimental approach presented in this work serves as a powerful framework to accelerate the discovery and development of oxidation-resistant HECs, paving the way for their expanded applications in demanding high-temperature, oxygen-rich environments.