As a big steel producer, China has a large number of steel consumption every year and it is still incredibly increasing now. The fast growth of steel consumption led to excessively production, which caused not only tons of material wastes but also seriously affect on economic and environment. Therefore, finding out and analyzing the driving force of steel consumption is an extremely importance research for reducing resource-consumption and pollutant-emission of steel industry. In this paper, the structure of steel consumption was established at first based on the bottom-up method. Further, the four factors closely related to steel consumption were identified by factor decomposition method. They are in-use stock of steel (Sn), average service life-span (Y), productivity per in-use stock of steel (H) and steel output per unit GDP (T). Next, the driving force of each factor was analyzed by their contribution rate to steel consumption of China. At the same time, the S-shaped model of growth, mathematical model and BP neural network model were used to extract four scenarios to simulate steel consumption. Results show that: (1) construction industry is the main industry of steel consumption, accounting for 50% of the total, and the rest of the industries is relatively low; (2) the driving force Sn of steel consumption gradually decreased while the rest factors (Y, H, T) on steel consumption increased every year within the research time. Especially, after 2009, the impact value of the four factors reached the same level of 25%; and (3) when the growth rate of GDP is keeping on 7%, steel output per unit GDP (T) will reach to 0.1248 t/104 RMB; the average service life-span of in-use stock of steel (Y) will increase to 8.4 years; and steel consumption will reach to 7.7×108 tons and the in-use stock of steel will be efficiently used.