In a two-period model, economists such as K.J. Arrow, A.C. Fisher, and C. Henry, have shown that when development is both indivisible and irreversible, a developer who ignores the possibility of obtaining new information about the outcome of such development will invariably underestimate the benefits of preservation and hence favor development. In this note, I extend the AFH analysis in two directions. I model the land development problem in a dynamic framework, explicitly specifying an information production function. In such a setting, I then ask and answer the question concerning when development should take place. Forthcoming in Journal of Environmental Management