Assessment of undiscovered oil and gas potentials can be made and presented in a probability format reflecting the inherent uncertainties and risks. A cumulative probability curve shows the chances of occurrence of possible hydrocarbon volumes, the risk that there is little or no potential, the average or expected value, and the "highside" potential. Such a curve can be drawn directly as a delphic consensus of expert opinions. Preferably, however, a curve can be constructed by multiplying, in a Monte Carlo simulation, several factors whose product is potential barrels of oil or cubic feet of gas. Each factor is entered as a range of values whose spread depends on the uncertainties. An example is the multiplication of the possible volumes of prospective sedimentary rock (e g., in cubic miles) by their potential hydrocarbon yields (e.g., barrels per cubic mile). Other geologic approaches depend mainly on estimated volumes of subsurface hydrocarbon trap space, on areal yields, on geologic analogies with other producing areas, on summation of individual prospect assessments, on the numbers and sizes of potential fields, or on the geochemical material balance of hydrocarbons generated, migrated, and trapped. More purely statistical methods involving the extrapolation of past discovery rates can be used only in maturely explored areas where data are abundant.