Accurately forecasting the customer baseline load (CBL) is of great significance for the incentive-based demand response in terms of determining the scheduling demand and quantifying the customer response. In recent years, with the gradual implementation of the dynamic time-of-use electricity price policy, the power consumption behavior of residential customers will become more complex and changeable. The impact of price fluctuation is not considered in the existing CBL forecasting methods, so there are large errors in the forecasting results. Therefore, a CBL forecasting method for residential customers suitable for dynamic time-of-use electricity price is proposed. The electricity consumption behavior of residential customers under the dynamic time-of-use electricity price is modeled as a household energy optimization management problem. By adjusting the weight between the two indicators of customer comfort and economy, the load data of customers with different preferences under the dynamic time-of-use electricity is simulated. Secondly, the impact of electricity price fluctuations on CBL is considered from the peak-valley attributes of the period to be forecasted and the electricity price difference between the current period and the adjacent period. The peak-valley attributes and price difference are extracted as input features together with the historical load to forecast residential CBL. The method proposed in this paper is compared with the traditional CBL forecasting method, and the results show that the proposed method can effectively improve the CBL forecasting effect under the dynamic time-of-use electricity price.