As a key component of transformers, oil-immersed paper insulation bushing can pose a threat to the operational safety of the transformer's main equipment if their insulation is damaged. The paper proposes a method for defect identification and insulation condition assessment of an oil-immersed paper insulation bushing based on multi-source fusion. First, typical defect test bushings are made. Then the types of bushing defects are identified by frequency domain dielectric response, partial discharge, and infrared thermal imaging related joint testing. Then the insulation index of the bushing is extracted and the membership function and state score operator are introduced. The evaluation model of bushing insulation state based on game theory combined weighting method is established. Finally, evaluation models using the combination and single weighting methods are used for example analysis of normal and defective bushings. The results show that joint testing can accurately identify the types of casing defects. The method avoids the problem of one-sided evaluation results using a single weighting method. The evaluation scores of normal bushings are higher than those of defective bushings. The correctness and effectiveness of the bushing insulation state evaluation model are verified.