In this paper, a second‐order cone programming (CP) method by which one can accurately calculate the interval power flow of an active distribution network with distributed generation is proposed. First, according to the interval models of the distributed wind power and solar power generation, the nonlinear programming (NLP) model for the interval power flow of the active distribution network is established. Second, based on the structure characteristics of the distribution network, the NLP model of the interval power flow is transformed into the standard second‐order CP model. Finally, an interior point method with polynomial time complexity is used to solve it. The numerical results obtained by the modified IEEE‐33 bus system show that the proposed second‐order CP method has better global convergence and a higher computational efficiency than the NLP method, and the interval for the power flow state of the distribution network can be accurately solved. © 2018 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.