The first part of the paper focusses on the implications of optimal intertemporal pollution control under alternative hypotheses about the assimilative capacity of environmental resources. Then it is shown how the optimal intertemporal allocation could be achieved by an emission tax policy. As is well-known, the informational requirements for implementing such an optimal pollution control are unsurmountable. In particular, there is no hope to get all the information about marginal individual valuations necessary for the fine tuning of "optimal" tax rates along the time path to the optimal steady state. Therefore the paper proceeds with investigating a number of environmental policy strategies "for the better" which are less demanding with respect to information on individual preferences, but which nevertheless use some information on individual marginal valuations and at the same time account for the ecological dynamics - to some extend, at least. It is shown that ignoring the ecological dynamics may lead to ill-defined policy options and to significant biases in measuring willingness to pay either with indirect or direct valuation methods.