Abstract: In recent years, a large number of distributed generations, electric vehicles, and energy storage on consumer side are connected to the distribution grid, which challenges the distribution grid operation. At the same time, the uncertainty of distributed renewable energy and the difference between electricity consumption of consumers make the three-phase distribution grid unbalanced, thus making the safe and economic operation of the distribution grid more difficult. However, voltage angle may not be available, which makes it difficult to identify the topology and line parameter. To address these problems, this paper proposed a data-driven method for the topology and parameter identification of three-phase distribution network without voltage angle. Firstly, the initial values of topology and parameters were obtained by using linear regression method without phase label; then, the phase sequence, topology and parameters were adjusted using correlation; finally, a specialized Newton Raphson method was proposed to obtain fine line parameters and recover voltage angle. The convergence and accuracy of the proposed method were verified by using IEEE 34 and 123-bus three-phase distribution feeders with real load data from Ireland.