A consistent, agreed-upon terminology is prerequisite for geothermal resource assessment. Accordingly, we propose a logical, sequential subdivision of the "geothermal resource base", accepting its definition as all the thermal energy in the earth's crust under a given area, measured from mean annual temperature. That part of the resource base which is shallow enough to be tapped by production drilling is termed the "accessible resource base", and it in turn is divided into "useful" and "residual" components. The useful component (i.e. the thermal energy that could reasonably be extracted at costs competitive with other forms of energy at some specified future time) is termed the "geothermal resource". This in turn is divided into "economic" and "subeconomic" components, based on conditions existing at the time of assessment. In the format of a McKelvey diagram, this logic defines the vertical axis (degree of economic feasibility). The horizontal axis (degree of geologic assurance) contains "identified" and "undiscovered" components. "Reserve" is then designated as the identified economic resource. All categories should be expressed in units of thermal energy, with resource and reserve figures calculated at wellhead, prior to the inevitable large losses inherent in any practical thermal use or in conversion to electricity. Methods for assessing geothermal resources can be grouped into 4 classes: (a) surface thermal flux, (b) volume, (c) planar fracture and (d) magmatic heat budget. The volume method appears to be most useful because (1) it is applicable to virtually any geologic environment, (2) the required parameters can in Sprinciple be measured or estimated, (3) the inevitable errors are in part compensated and (4) the major uncertainties (recoverability and resupply) are amenable to resolution in the foreseeable future. The major weakness in all the methods rests in the estimation of how much of the accessible resource base can be extracted at some time in the future. In a manner similar to mineral and fuel assessment, this recoverability is expressed as a "recovery factor". For an ideally permeable hot-water system, the recovery factor may be as much as 50% and seems to be independent of temperature. It must decrease as effective porosity (??e) decreases, but the relation between the two is little more than a guess. On the other hand, for favorable systems like Larderello that produce steam by a mechanism of intergranular vaporization, the recovery factor is probably around 15-20%, decreasing to zero at an effective porosity of zero. According to the anlysis of Bodvarsson (1974), it increases with decreasing reservoir temperature, and as pointed out by Nathenson (1975a) is limited at low temperatures by the need to have sufficient reservoir pressure for extraction and use. The extent to which a geothermal reservoir can be resupplied with heat during "industrial" times of 10-100 yr can be evaluated using simple analytical models. The results, combined with gravity and levelling data of Hunt (1977) for Wairakei and Isherwood (1977) for The Geysers, confirm earlier conclusions by Ramey (1970) and Nathenson (1975a) that resupply to reservoirs producing only steam can be neglected, and the conclusion of Nathenson (1975a) that it may be significant for hot-water systems of high natural discharge. Major subjects that demand continuing investigation include: 1. 1. Determination of recovery factors as functions of temperature and effective porosity, particularly for hot-water systems. 2. 2. Evaluation of fluid recharge and heat resupply by repetitive gravity, levelling and underground temperature surveys in producing geothermal fields. 3. 3. Analysis of the extent to which a recovery factor can be enhanced by stimulation and by use of confined circulation loops. ?? 1979.