Facing the tense situation of the world's energy, it is particularly important to develop and utilize environmentally friendly, renewable energy grass specially prepared for cellulosic ethanol production. Therefore, the establishment of a whole process of energy grass life cycle assessment model of production of cellulosic ethanol is necessary. This paper gives an inventory analysis of the life cycle of the process that uses energy grass to produce cellulosic ethanol, according to the theory of life cycle; and then uses BP neural network to assess environmental impact of this process. Research results show that the BP neural network meets the requirements of this environmental impact assessment, atmospheric pollutants and water pollutants of energy grass of cellulosic ethanol production process have reached the level of standard.