Revenue management on intemet banner advertising based on CPM(cost per thousand impression)is studied in the thesis．Capacity control is an important technique of revenue management which is applied for managing perishable assets．Existing problems of current internet banner ads running status are put forward through analyzing the characteristics of internet banner advertising,The thesis adopts capacity conlrol to solve the sales problems of intemet banner advertising in order to increase advertising revenue for intemet advertising service Providers. As the limitation of webpage impression,capacity control policy can be studied to help the service provider allocate the intemet advertising resources efficiently． Introduction to Intemet Advertising is given in Chapter 1．The advantages and differences of internet advertising from traditional advertising are introduced briefly．In research background,the state of major portals in China in online advertising business is expatiated．The current operation of intemet banner advertising is analyzed.The state of intemet display related advertising including its pricing model is introduced．Based on the previous studies mentioned above,existing problems which will be studied in the following chapters are put forward. Chapter 2 reviewa the previous studies on capacity control，revenue management in traditional broadcasting industry,research on intemet advertisements including scheduling problem，pricing model and consumers’attitudes on intemet ads．As intemet advertising is a perishable commodity,revenue management can be applied to the research on the sales issue． Capacity control of single product is studied in Chapter 3．In this research, only one broadcasting day is taken into consideration．A dynamic programming model of capacity control is given to describe the issue．The property of value function is proven and dvnamic programming is adopted to solve the model．Finally,an example is given to analyze the sale policy．The result of the example proves the property of value function again and provides price policy at any remaining capacity and selling time．The protection level is given according to the policy of capacity control． Capacity control of two-product category is researched in Chapter 4．Advertisers can purchase one arbilrary category or purchase both of the categories．Similarly,a dynamic progvanmaing model of capacity control is given to describe the issue．Dynamic programming is adopted to solve the model．Then a numerical example is studied to analyze the sale policy．The result of the example shows that the property of two-category model is different from the single product．Moreover,the capacity control policy can’t be expressed by protection level．Therefore．The control policy of two-product is not as clear as single product case However,the capacity control can also be discussed according to the acceptable price set. Chapter 5 is conclusions and future searches. This chapter concludes the contributions brought the research and limitation of the thesis. The future researches in this field are suggested in this chapter.