There are 20 known high-temperature geothermal areas in Iceland and another eight potential areas. Surface manifestations are meagre in these eight areas and not conclusive, and no drilling has been carried out to prove or disprove the existence of high-temperature geothermal systems at depth. The high-temperature areas are located within the active volcanic belts or marginal to them. The heat source is considered to be magmatic, shallow level crustal magma chambers in the case of high-temperature systems associated with central volcanic complexes, but dyke swarms for the systems on the Reykjanes Peninsula where no central volcanoes have developed. Fossil high-temperature systems are abundant in Quaternary and Tertiary formations as witnessed by alteration of the basaltic eruptive rocks into lower-greenschist mineral assemblages. The fossil systems are typically associated with central volcanoes where intrusives account for 50% or more of the rock. The fossil systems are considered to have formed within the active volcanic belts but drifted out of them in conjunction with crustal accretion within these belts. In the process they may develop into low-temperature geothermal systems. Permeability is very variable within the drilled high-temperature areas, in the range 1–150 millidarcies. The best permeability generally appears to be associated with sub-vertical fractures and faults. Permeability is poorest when the reservoir rock consists dominantly of intrusives, such as at Krafla, northeastern Iceland. It appears that intrusives are most abundant in reservoirs associated with central complexes that have developed a caldera. Temperatures follow the boiling point curve with depth, at least to the level of the deepest wells, in some areas, but in others they are lower. The highest recorded downhole temperature is >380°C. Hydrological considerations and permeability data favour that convection is density driven and that the source water is shallow groundwater in the vicinity of these systems. This groundwater is in most cases of meteoric origin. However, in three areas on the Reykjanes Peninsula it is largely or solely marine. The deuterium content of geothermal waters of meteoric origin is often lower than that of local precipitation. This has been taken to indicate that the source of supply is precipitation that has fallen on higher ground inland. This may indeed be the case, but flow from the source area is considered to be shallow. In some cases the low δD-values may stem from the presence of a component of an old water, which is isotopically lighter than today's precipitation at any particular site because the climate in Iceland was colder in the past. The geothermal seawater at Reykjanes and Svartsengi, southwestern Iceland, is considerably lower in deuterium than seawater. The cause of this is not known. However, reaction between seawater and basaltic rocks at very low temperatures may contribute, as well as rising of H 2 gas from deep levels and its reaction at shallower levels in the geothermal system to form water, but H 2 gas is much more depleted in deuterium than the associated water. Degassing of the magma heat source appears to add chemical constituents to the geothermal waters, such as boron, carbon and sulphur. Sometimes there may also be addition of Cl and H 2O during events of recharge of new magma into the magma chambers in the roots of the geothermal system such as has been observed in the Krafla area. The high-temperature geothermal waters are close to chemical equilibrium with alteration minerals for all major components, except Cl and B. The alteration minerals typically display depth zoning because many of them are stable only over a limited temperature range. At temperatures above about 250°C the alteration mineral assemblage is that of the greenschist metamorphic facies. Precipitation of carbon as calcite and sulphur as sulphides, where boiling occurs in upflow zones of high-temperature geothermal systems, leads to strong enrichment of carbon and sulphur in the altered rock.