For the time being, the condition assessment of transformer doesn’t take the variation of multi-stage information into consideration and the weight determination of assessment index is relatively simple, causing the overall assessment bias. Against that, this paper proposes a transformer condition assessment model based on optimal weight and dynamic grey target with interval grey number. First of all, it selects the status assessment indicator and classifies the index data into efficiency index and cost index and then makes normalized treatment. Secondly, AHP is used to assign the subjective weights of the indicators; the variance difference and the mean difference are introduced to measure the volatility of the interval index data and the objective weight is given. The optimal weight of the index is determined by establishing the optimal model based on relative entropy principle. Next, an assessment method based on dynamic grey target evaluation model of interval grey number is proposed considering the accumulation of transformer phase information and the dynamic change of index, which seeks approaching degree to evaluate the state of transformers. Finally, the multi-stage historical data of multiple transformers in a substation is analyzed, and the objectivity and validity of the method are verified. This work is supported by National Natural Science Foundation of China (No. 51777119).