Expert systems, though not widely used in the oil industry, have been the object of a large volume of research and development activities in the industry and in academia. Two reasons for limited practical usage of expert systems in oil exploration are (1) exploration in general is highly multidisciplinary, and (2) rules governing the exploration process are, for the most part, subjective. The combination of these two factors has made development of expert systems for solving practical exploration problems difficult. Recent advances in some areas of expert systems, coupled with the availability of cost effective and fast workstations, offer opportunities to overcome the two major obstacles. Specifically, using concepts such as evidential reasoning, fuzzy logic, and neural networks in expert systems makes integration of different knowledge sources, implementation of inexact and qualitative rules (information), and self learning more practical.