Many web sites provide free services to users. The revenue for these sites is dependent on the advertisements they place on their web pages. For these firms, therefore, the optimal placing of advertisements on their web pages becomes a critical issue. In a given planning horizon, we assume that the number of advertisements available for placement on a web site is greater than the available space. The site, therefore, needs to determine the optimal allocation of advertisement space. We develop a hybrid genetic algorithm (GA) that uses problem specific knowledge during the evolution of solutions to solve this problem. Our initial computational results show that the hybrid GA performs exceptionally well in the sense that it provides optimal or near optimal solutions for a variety of problems.