According to the construction of infrastructure for the water resource management services, the importance of the eco-friendly reservoir tank management is being spotlighted. In this paper, we proposed the eco-friendly reservoir tank management using prediction for improving the water quality and on-line managing efforts of reservoir tanks. The proposed method defined the context and environment of the reservoir tank and predicted the profited service according to the pump motion, the solar battery, the chemicals, the water level, the telephone line, and the modem using collaborative filtering. To evaluate the performance of the eco-friendly reservoir tank management system using prediction, we conducted sample T-tests so as to verify usefulness. This evaluation found that the difference of satisfaction by service was statistically meaningful, and showed high satisfaction. Accordingly, the satisfaction and the quality of services will be improved the efficient prediction by supporting the context information as well as the environment information.