The function of thyroid glands of ground squirrels, maintained under conditions of constant illumination and temperature as measured by the release of radioiodine (125I), varied from one time of the year to the next and differed among species. The maximum release rate was 2.476% per day and was observed in the high altitude species, Spermophilus lateralis, while the slowest release rate was 0.029% per day and was observed in the desert species, S. townsendi. The periods of slowest release tended to coincide with late summer and early fall, though individual animals of the eight species in which 125I release rates were studied varied considerably among themselves. The variation within species was thought to reflect the free-running circannual rhythm of thyroid secretion which eventually became out of phase with natural conditions. The reason for differences among species is unknown, though it may be related to water balance. The rhythm of thyroid activity is not a function of body temperature rhythms, but reflects primary rhythms such as TSH (thyroid-stimulating hormone) secretion or release of TSH. Although the time of year when the thyroids are least active overlaps the period of hibernation, thyroid inactivity cannot be a causal factor, since exogenous thyroxine does not prevent torpor and the period of reduced thyroid activity is not precisely correlated with hibernation. Increased thyroid activity typically occurs during the winter and early spring, times when animals may still be in hibernation, at least in nature. This is interpreted to mean that their glands become active during the hibernation period in preparation for reproduction and molting. All of the nine species of ground squirrels in which body temperature and metabolism were measured had basal metabolic rates which could be described by the equation: cm3O2 · g-1 · h-1 = 3.24 W-0.23, although the only nonhibernator in the group, Ammospermophilus leucurus, had a metabolic rate 4%-13% above the level expected from this equation. Removal of the thyroid gland did not change the metabolism of S. mexicanus, while it lowered the metabolism of A. leucurus and S. townsendi 13%-15%, which is far less than expected for animals such as the white rat, in which the thyroid gland contributes a significant amount of thermogenesis. This is consistent with the hypothesis that ground squirrels in general have a low metabolism due to a relatively inactive thyroid. The effect of thyroidectomy on metabolism is related to the time of year in which the gland is removed, so that removal during periods of slow release rate of 125I does not lead to a reduction in metabolism. In general, metabolism declined seasonally independent of whether the thyroid was present or absent. The seasonal changes in metabolism are not due to changes in body weight. In general, body temperature tended to decline when metabolism declined and vice versa.