Most of ecosystem services, such as water service and climate regulation et al, are public good. Consequently, the value of these ecosystem services are not embodied in the prices of conventional products, and therefore the market does not provide people whit the economics incentives to supply the amount of ecosystem services demand by society. Payment for ecosystem services (PES) is emerging as a new approach to managing the valuable services derived from ecosystem globally. It provides farmers with incentives to increase the supply of ecosystem services, so it tends to be easier to secure cooperation from land users and protect the ecosystem. When come to practice, how the level of payments is related to the supply of ecosystem service is a very important problem. Traditionally, it needs to integrate physical and economic models at a disaggregate level necessary to capture the heterogeneity of the physical environment and the economic behaviors of landholders'. However, high-resolution biophysical and economic data with the geographic coverage needed for analysis are exceptional. In most cases, the time and resources required to get these data precludes their use for most policy analysis. In order to provide timely, sufficiently accurate information to support policy decision making, the minimum-data (MD) approach was introduced, which is a trade-off between timeliness and accuracy of information. The MD approach exploits the structure of the PES problem to obtain an approximation to the ecosystem service supply curve that can be implemented using relatively simple data, This method shows that the supply of ecosystem services can be derived from the spatial distribution of opportunity cost of providing those services on the assumption that landholders take land use and management decision to maximize their perceived economic well being. Then the MD approach was applied to simulate the supply of water services from agriculture in Ganzhou District, which located in middle reaches of Heihe River. The water services from this area are measured by the difference of evapotranspiration between two kinds of land-use: irrigated land and non-irrigated land. The spatial distribution of opportunity cost of providing water services was obtained by undertaken a field survey (n=120). Under some plausible simplifying assumptions, the supply curve of water services was approximately determined. The results show, as the price of payments increasing, more land chooses to take the land-using which in favor the water services supply. When the price increase from 0.067 RMB m-3（irrigation water price）to 3 RMB m-3 (construction water price), the additional added water services are increase from 2.08×106m3 a-1 to 2.7×108m3 a-1. This means that the payments for ecosystem services can be used to decrease the water consume of agriculture, and increase the water supply for the ecosystem which would better off the environmental condition of downstream. The MD appears to be a transparent and cost effective tool to quantify the effect of financial incentives in the provision of water resources. Policy relevant information can be generated without the need to conduct expensive field surveys and to set up more elaborate econometric simulation models. Also, by using the MD, the appropriate level of price under certain objective of ecosystem services was determined.