The degradation of grassland in China is serious, the area of degradation accounting for 90% available grassland in China. Serious grassland degradation has brought great impact on local ecological environment, biodiversity conservation and animal husbandry development. In Taipusi Banner, Inner Mongolia, severe over-grazing resulted in the area of degradation accounting for 90% of the 150 000 hm2 of typical steppe. To evaluate the public grazing management of typical steppe in this region, the condition of grassland, animal, farm income, climate and some control experiment data were used in this bioeconomic model of China to analyse public grazing system in Taipusi Banner. These analyses were used two farm models. One was feeding balance model which was used to analyse the livestock production system and the second was farm systems model that which was used to analyse the optimal solution for net financial returns within specified constraints. The farm surveys results showed the current practice of high stocking rates in Taipusi directly leads to grassland degradation, low animal productivity and net incomes. Simulations of the feed balance model indicated that there were big energy demand deficit from October to May at animal growth time, net income of farm was inclined to decreased when stocking rate was more than 2.275sheep/hm2. The farm systems model showed that net financial returns could be improved through reducing stocking rates closer to 2.275sheep/hm2 for the current predominately mutton sheep system. And further gains could result from changing to fine wool sheep at the same lower stocking rate. At the same time, reduce goat number could enhance the grassland productivity, especially when combine with an early summer grazing ban that enables the grassland to re-establish after winter. Estimates of grass condition demonstrated that a different ratio of desirable and undesirable grassland species has no significant effect on net income and feeding cost when stocking rate is less than 1.55sheep/hm2. A higher composition of desirable species resulted in lower feeding costs and greater net income when the stocking rate was more than1.55sheep/hm2. Net income can be maintained at the highest stocking rate when desirable species exceeded 75% of the grassland composition. Consequently, the optimal stocking rate of Taipusi Banner is between 1.55-2.275sheep/hm2. Sustainable grassland and livestock production requires a systems analysis of current farm practices and options. The analysis of a typical Taipusi farm model showed that individual farmer can adopt grazing management changes were realistic and potentially. Importantly, the potential management options are compatible with current farming systems. The directions of change are useful for achieving benefits for grassland productivity improvement, though further research will be needed to proved whether rate of change is sustainable or not. Our study tested that improvement productivity of farm and net farm financial returns could be achieved at the same time by reducing stocking rates, improving supplement condition, improving desirable grass ration and changing enterprise types. Our study was expected to provide meaningful data for preventing grassland from degradation, restoring grassland, grassland optimal utilization, improving grazing management and livestock production as well.