The typical theoretical work on dividend policy suggests five possible imperfections that management should consider. They are taxes, asymmetric information, an incomplete contract, institutional constraints and transaction costs. Different from the typical framework, this dissertation is to study the optimal dividend poly with heterogeneous beliefs among investors. The first model in this study has analyzed investment/dividend policy with heterogeneous beliefs-the full information model in a frictionless economy with divergent types of shareholders. A high dividend policy is optimal with limited endowment for the optimistic investors as the stocks are sold not only to type-o investors, but also to at least one type-p investor holding some shares. A low dividend policy is appropriate with cash dividend D= X0-ao+1 is optimal as the shares are sold only to the type-o investors. Heterogeneous beliefs of investors change dividend policy given the same information even under full information. Following the Miller and Rock (1985) theory, the second model in this dissertation has analyzed heterogeneous beliefs among investors-the two period model in leading to changing a firm¡¦s optimal dividend policy. A firm¡¦s optimal dividend policy is changed not only by the ratio of the pessimistic to optimistic investors, but also heterogeneous beliefs. An increase in the ratio of pessimistic to optimistic investors will result in a higher dividend. On the other hand, as the beliefs of both optimistic and pessimistic investors increase, i.e. a new biotechnology is innovated, a relative low dividend policy is appropriate. Based on the previous analysis, the results show that optimal dividend policy with heterogeneous beliefs among investors in a firm¡¦s earnings exists under heterogeneous beliefs framework. A firm¡¦s optimal dividend policy is different from that of the MM dividend invariance theorem. It is not because of taxes, asymmetric information, incomplete contracts, institutional constraints and traction costs, but heterogeneous beliefs of investors.