The loss of natural habitat from conversion to human dominated uses is the major cause of the decline of terrestrial biodiversity. The formation of networks of natural reserves is a cornerstone conservation strategy, but existing reserve networks are nowhere near what is necessary to protect existing biodiversity. Much of the existing literature on systematic conservation planning is within a static context even though both conservation planning and habitat loss via development are ongoing processes which unfold over time. Some areas face imminent threat where if action is to be taken to protect biodiversity, it must be done with urgency while other areas face much lower threats where conservation actions could be postponed for decades with little risk of biodiversity loss. Thus, an important question is not only where to take action, but also when. Using ecoregions as the basic unit of analysis, this project develops and applies a dynamic model of conservation planning. In each ecoregion, species-area curves determine biodiversity conservation as a function of remaining habitat area. The model incorporates the initial amount of protected area, the increase in protected area as chosen by a conservation planner, the initial amount of developed area, and the rate of habitat conversion from development. At every instant in time, a conservation planner chooses where to allocate conservation resources for the purposes of establishing (expanding) biological reserves and preventing habitat from being lost to development. The model solves for the optimal allocation of a conservation budget across ecoregions through time. Optimal strategies involve equalizing marginal gains per dollar across all ecoregions if at an interior solution where no conservation choices have been limited by the scarcity of available land. Corner solutions, where available land has been used up before the desired level of conservation have been achieved in one or more ecoregions require more complex strategies, namely first allocating to ecoregions where the constraint is most binding until exhausting available land and then switching to other less pressing ecoregions. This dynamic model is applied to terrestrial ecoregions of the continental U.S. to solve for optimal allocations of a conservation budget over time. A variety of biodiversity metrics are considered, including species richness, species endemism and phylogenetic diversity. Various extensions are considered, including comparison of optimal allocation strategies to more tractable heuristics, comparison of optimal strategies to specified targets by specified times (resembling actual current conservation planning agency goals), endogenous land prices as a function of land scarcity, and multiple conservation strategies besides reserves creation.