This thesis investigates how the extraction of natural resources is affected by expropriation risk. Conventional Hotelling-type theory suggests that extraction rates will increase when companies are faced with expropriation risk, because the value of resources left in the ground decreases. However, when capital investments are included in the framework there is a second effect of expropriation risk, which depresses investments in physical capital, leading to lower extraction rates. The total effect is theoretically ambiguous, and arguably depends on the capital intensity of the resource. I show that iron, bituminous coal and non-bituminous coal can be ordered in terms of capital intensity, from high to low, and use this information in a pooled OLS regression to show that the effect of expropriation risk is more negative the more capital-intensive the resource is. The effects are quite large: A one-standard deviation increase in the expropriation risk index is associated with 44 %, 37 % and 24 % decreases in the extraction rates of iron, bituminous coal and non-bituminous coal, respectively. These results suggest that the quality of property rights institutions is important for efficient exploitation of a country's natural resources.