Overcapacity is a persistent problem in China's economy. Previous de-capacity measures based on the quantity of the total excess capacity have failed to prevent the recurrence of increasingly worse overcapacity. To solve this problem we attempt to quantitatively analyze China's excess coal capacity by dividing it into the long-term natural excess capacity and short-term cyclical excess capacity using the state-space model and Kalman filter algorithm. The results show that China's excess coal capacity can indeed be divided into natural excess capacity and cyclical excess capacity, and they have different causes, fluctuations, roles, and effects on the price. In 1995-2001, cyclical excess capacity was the main factor of overcapacity, but in 2002-2015 natural excess capacity played a key role. Cyclical excess capacity has a negative effect on the price, whereas natural excess capacity has little effect on it. Therefore, policy makers should focus on the causes, the fluctuations, and the roles of natural excess and cyclical excess in addition to those of the total excess capacity. They should also consider the relationship between the capacity and price based on the different effects of the natural excess capacity and cyclical excess capacity on the price.