Abstract: Currently, the wind power development is mainly constrained by the phenomenon of wind curtailment. Also, the traditional uncertainty modeling methods（e.g. stochastic programming and robust optimization） may result in conservative or uneconomic decisions in dealing with the wind power uncertainties. Based on challenges mentioned above, a distributional robust coordinated dispatch model for integrated electricity and heating system considering uncertainty of wind power was proposed in this paper. First, a deterministic coordinated dispatch model of the integrated system was presented, which regarded generation cost for traditional units and combined heating power（CHP） units and curtailment cost of wind power as its optimization target, and contained the constraints such as power balance, heating balance, minimum ON/OFF time. Then, on the basis of the existing huge historical data in the SCADA/EMS system, a data-driven based two-stage distributional robust dispatch model was formulated with wind power uncertainty considered. In the first stage of the two-stage model, the objective function included not only the start-up and shut-down cost of the traditional generation units but also the corresponding cost associated with the forecast wind power scenario. This treatment method can help to obtain a more economical dispatch planning strategy. In the second stage, norm-1 and norm-inf were simultaneously included to constrain the confidence set of wind power probability distribution in order to find the optimal solution under the worst probability distribution. And, the distributional robust model was solved by the Column-and-Constraint Generation algorithm. Finally, the numerical results on the IEEE 39 bus system verify the effectiveness of the proposed model.