Load density index method is one of widely used spatial load forecasting techniques in distribution planning. The determination of load density is a key task for this method. In order to overcome the drawback of traditional method which adopts analogy method to obtain load density and is hard to meet the demand of precision, a new approach based on intuitionistic fuzzy entropy and fuzzy clustering is proposed. The fuzzy clustering analysis is adopted to classify the set of load density and impacting factors into several grades. Then the intuitionistic fuzzy theory is utilized to describe the uncertainty that would appear in the load density selection. According to the intuitionistic fuzzy entropy the grade of load density can be selected, and the load density interval is available by the fuzzy clustering from the grade selected. An example is used to demonstrate the validity of the proposed method.