Abstract Concentrations of the rare earth elements (REE), Th and U have been determined in thermal waters emerging from a number of locations in and around the Idaho Batholith. Previous investigators have suggested that the source of heat for the geothermal systems studied is the radioactive decay of K, Th and U which are enriched in the rocks through which the fluids flow. Thus, knowledge of the behavior of REE, Th and U in these systems may contribute to a better understanding of the potential consequences of the interaction of hydrothermal fluids with deeply buried nuclear waste. Such studies may also lead to the possible use of REE as an exploration tool for geothermal resources. The thermal waters investigated may be characterized as near-neutral to slightly alkaline, dilute, NaHCO 3-dominated waters with relatively low temperatures of last equilibration with their reservoir rocks (<200°C). REE, Th and U concentrations were measured using Fe(OH) 3 coprecipitation, followed by ICP-MS, which yielded detection limits of 0.01–0.003 μg/l for each element, depending on the volume of fluid sample taken. The concentrations of REE, Th and U measured (from <0.1 up to a few μg/l) are 3–5 orders of magnitude less than chondritic, in agreement with concentrations of these elements measured in other similar continental geothermal systems. The REE exhibit light REE-enriched patterns when normalized to chondrite, but when normalized to NASC or local granites, they exhibit flat or slightly heavy REE-enriched trends. These findings indicate that the REE are either taken up in proportion to their relative concentrations in the source rocks, or that the heavy REE are preferentially mobilized. Concentrations of REE and Th are often higher in unfiltered, compared to filtered samples, indicating an important contribution of suspended particulates, whereas U is apparently truly dissolved. In some of the hot springs the REE concentrations exhibit marked temporal variations, which are greater than the variations observed in major element concentrations, alkalinity and temperature. There are also variations in the fluid concentrations of REE, Th and U related to general location within the study area which may be reflective of variations in the concentrations of these elements in the reservoir rocks at depth. Thermal waters in the southern and central parts of the field area all contain ∑REE concentrations exceeding 0.1 μg/l (up to as high as 3 μg/l), Th exceeding 0.2 μg/l and U generally <0.4 μg/l. In contrast, thermal waters from the northern area contain lower ∑REE (<0.6 μg/l) and Th (<0.1 μg/l), but higher U (>3.0 μg/l). Using experimentally measured and theoretically estimated thermodynamic data, the distribution of species for La, Ce and Nd have been calculated and also the solubility of pure, endmember (La, Ce, Nd) phosphate phases of the monazite structure in selected hot spring fluids. These calculations indicate that, at the emergence temperatures, CO 2− 3 and OH − complexes of the REE are the predominant species in the thermal waters, whereas at the deep-aquifer temperatures, OH complexes predominate. In these thermal waters, monazite solubility is strongly prograde with respect to temperature, with solubility often decreasing several orders of magnitude upon cooling from the deep-aquifer to the emergence temperature. At the surface temperature, calculated monazite solubilities are, within the uncertainty of the thermodynamic data, comparable to the REE concentrations measured in the filtered samples, whereas at the deep-aquifer temperature, monazite solubilities are generally several orders of magnitude higher than the REE concentrations measured in the filtered or unfiltered samples. Therefore, a tentative model is suggested in which the thermal fluids become saturated with respect to a monazite-like phase (or perhaps an amorphous or hydrated phosphate) upon ascent and cooling, followed by subsequent precipitation of that phase. The temporal variations in REE content can then be explained as a result of sampling variable mixtures of particulate matter and fluid and/or variable degrees of attainment of equilibrium between fluid and solid phosphate.