Three simple methods for geothermal resource estimation—power density, stored heat, and exergy—are reviewed and their shortcomings are discussed. The conclusion is that computer modeling should be used instead of these simple methods. An overview is presented of the theory of and numerical methods used in computer modeling of geothermal systems. The geothermal modeling process of conceptual modeling, natural state simulations, matching production history, and future scenario simulations is discussed, including model design and boundary conditions. The problem of model calibration using both manual methods and inverse modeling software, such as iTOUGH2 and PEST, is discussed. Special topics in modeling engineered geothermal systems, including chemistry in geothermal models, are briefly reviewed. Finally, current development and future directions in geothermal modeling are considered, particularly the recent use of uncertainty quantification.