In this paper, we study the following question: For a public good economy where the provision of public goods is to be financed by property taxes collected from individuals, what is the optimal feasible tax mechanism when a social planner is relatively uninformed of the properties of the individuals? Using a Bayesian model, we provide the full characterization of the optimal feasible tax mechanism with two agents and its properties. We find that (i) when the expected total endowment of the economy is relatively low enough or high enough, the incentive compatibility constraint does not bind so that first best taxation can be obtained; (ii) the second best feasible tax mechanism requires a poor agent to pay relatively more than a rich agent, that is, it is regressive; and (iii) the optimal feasible tax mechanism is increasing in the sense that the agent's tax payment increases with his endowment. For the case of more than two agents, under certain mild assumptions we give some partial results similar to (i) and (ii) above. In addition, we find the optimal feasible tax mechanism for the corresponding infinitely large economy.