Inter-regional electricity transmission inevitably brings about unbalanced carbon emission and air pollution among different regions. It is vital to compensate for the economic and environmental loss of power output region by taking the carbon subsidy strategy into account. In this study, a risk explicit interval multistage stochastic programming model considering inter-regional carbon subsidy is developed to support an optimal energy system planning in Inner Mongolia, China. This model can not only deal with uncertainties embedded in the complex energy system described as interval values and random variables but also reflect dynamic linkage between multiple stages over a time series. It advanced the existing optimization methods by introducing the cost-risk tradeoff information based on the risk preferences of decision-makers. The obtained crisp solutions are more feasible, effective and optimum for the practical policymaking process. Different scenarios associated with carbon subsidy policy implementation, renewable energy share adjustment and carbon emission reduction strategy are employed to project the optimal power generation scheme, net system cost, installed electricity capacity, carbon intensity and pollution emission in the future. The results indicate that the amount of exported electricity will keep increasing, and the carbon subsidy strategy will prominently promote the economic growth of the electric power sector by 3.05%. The further power supply pattern is going to transform from fossil fuel to renewable energy, by which wind and solar power will occupy around 44% of total power generation capacity in West Inner Mongolia. Raising the share of renewable energy will be the most effective approach to adjust the power generation scheme; meanwhile, reducing carbon intensity will significantly contribute to enlarge green energy capacity, enhance low-carbon emission and mitigate air pollution. These findings will help decision-makers gain insights into a more scientific energy system planning under various uncertainties.