In a grid computing environment, resources are autonomous, wide-area distributed, and what's more, they are usually not free. These unique characteristics make scheduling in a self-sustainable and market-like grid highly challenging. The goal of our work is to build such a global computational grid that every participant has enough incentive to stay and play in it. There are two parties in the grid: resource consumers and resource providers. Thus the performance objective of scheduling is two-fold: for consumers, high successful execution rate of jobs, and for providers, fair allocation of benefits. We propose an incentive-based grid scheduling, GridIS, which is composed of a P2P decentralized scheduling framework and incentive-based scheduling algorithms. Simulation results show that GridIS guarantees the incentive of every participant to a satisfying extent.