This research project develops and extends a Hotelling type model of the optimal extraction of a nonrenewable natural resource. From the model, the relationships between the resource price movements and such factors as discount rates, market structures, extraction costs and resource reserves are obtained. The major theoretical predictions gained from the model are 1) with endogenous discount rates, as discount rates are high, a natural resource would be depleted more slowly; 2) a monopolistic extraction of a natural resource depends not only on discount rates, but also on demand elasticities compared with a competitive industry, and 3) the traditional increasing price paths of exhaustible natural resources could be altered by the rate of changes in marginal extraction costs. The model suggests a declining and a U-shaped time path for prices with a decreasing rate of change of marginal extraction costs. This research project also empirically tests the price behavior for ten different mineral commodities based on annual data over the period 1950:1-2010:12. The empirical study indicates that substantial differences across different natural resource markets make it difficult.