Abstract: Large-scale fracturing technique is one of the important means to effectively develop shale oil. After large-scale fracturing, complex fracture networks will be formed around the wellbore. The inversion of fracture parameters after fracturing is the key to fracturing performance evaluation and parameter optimization for hydrocarbon development. However, it is difficult for the existing mathematic and well test models to meet the inversion needs of complex fracture network in shale oil. For this reason, the paper investigates the characterization of complex fracture networks in shale oil fractured horizontal wells, establishes semi-analytical well test models for multi-mode fracture network of shale oil, including volume fracturing model, compound fracturing model, and discrete fracture network model. The semi-analytical well test model of multi-mode fracture network is solved using point source method, semi-analytical method and Laplace transformation. Moreover, efforts are made to perform numerical verification, divide flow phases and analyze the characteristics of flow section. Based on the semi-analytical well test model established for multi-mode fracture network, a sensitivity analysis was performed on well test characteristic curve, and then a well test curve fitting method was established. Supported by the production history fitting method, the scheme for parameter evaluation of shale oil multi-mode fracture network was initially developed based on well test theory. The fracture network parameter evaluation method was actually applied in the shale oil fractured horizontal wells JA and JB in Jimsar. An evaluation was performed on fracture network parameters, including fracture network geometries, half length of primary and secondary fractures, hydraulic conductivities of primary and secondary fractures, storage coefficient before and after crack closure, and fracture closure time. In addition, the accuracy and practicality of the evaluation method of complex fracture network parameters were proved by field applications.