Abstract: A real-time divisible loads scheduling problem with the service benefit maximized and the make span of the loads minimized in heterogeneous distributed systems was investigated. For the sake of increasing the profit of service and maintaining a high ratio of acceptance, three efficient strategies were proposed:(1) An acceptation/rejection judgments strategy, which takes service benefit and make span into consideration. (2) A strategy of deleting the loads from the load waiting queue; (3) A sorting method of least laxity first strategy with profit of service considered. To minimize the make span, a scheduling algorithm was designed, which can use the free time sufficiently. Experimental results demonstrated that the proposed algorithm can obtain a higher gain for the service provider with the higher acceptance ratio and the shorter make span.