Abstract: In view of the fact that the determination of variable weights has not used fully the acquired forecasting data, a method of multi-step variable weight combination forecasting is proposed to forecast China's postal service revenue and postal staff from 2005 to 2010 as an example in application. The method is used to determine the weights in each period by making full use of acquired forecasting data together with the multi-step solution to quadratic programming, then the weights are reused for combination forecasting. The forecasting results show that the method is rational and easy to operate. China's postal service revenue will increase gradually in the years to come. However, the number of postal staff will decrease because of the improvement of labor productivity and reduction of redundant staff, but the reduction tends to slow down. On the other hand, it should not be overlooked that the postal service can provide jobs for employment and is necessary for national economic growth.