These studies of the influence of environmental temperature on the movements of tadpoles of Bufo terrestris americanus were part of an investigation of tadpole aggregational behavior in northern Michigan during the summers of 1963-64. Fieldwork was done at a pond and surrounding small pools in a shallow sand pit. Temperatures and larval distribution were recorded along a transect extending 6 m from shore toward the center of the pond. As the water warmed in the sun, the tadpoles left their locations scattered throughout the pond and moved up the temperature gradient, collecting in shallow water, where the maximum temperature found was 31.2 deg. In daytime most of the larvae were in the upper 1-2 deg of the temperature gradient. In the evening they dispersed rather uniformly throughout the lowered temperatures of the entire pond. Laboratory experiments showed similar behavior by samples of 75-100 tadpoles in gradients of temperature artificially developed in a shallow trough 1.5 m long. Usually the animals moved up the gradient all along the trough, most of them stopping in the highest degree or two. As the temperature there rose above 31 deg., larvae began to move away, and all had left at 39 deg. Lethal limits for individual tadpoles varied from 37.2 to 42.0 deg. Temperatures in small isolated pools beside the pond reached lethal range, as high as 37.8 deg. From an early morning location in the deep part of the pool, the tadpoles moved into the shallow margins as the water warmed and developed a gradient of about 1 deg. When the entire pool was 30-30.5 deg. they distributed themselves nearly uniformly throughout it. At the maximum levels (36.3-37.8 deg. in different pools) they collected in the deepest part, putting the thickest possible layer of radiation-absorbing water above them. At these times the gradient was less than 1 deg. As temperatures receded, the tadpoles again collected in the shallows, even though the gradient was only 0.5 deg. The similarity found between the maximum field temperature (37.8) and the minimum lethal temperature (37.2) suggests a potent selective role for water temperature in small pools. Tadpoles killed in these circumstances probably are very important nutritionally to the remaining larvae in these food-deficient conditions. Dense aggregations comprising a minority of the larvae, were present in small pools at all temperature levels, though in one pool they dispersed at 37 deg. None was formed in the experimental gradients. Thus our results include no indication that their formation is controlled by temperature, as has been suggested. Although tadpoles are doubtless heated directly by the sun, our data do not support the hypothesis that groups radiate enough heat to the surrounding water to create temperature gradients sufficient to attract other tadpoles to form dense aggregations. There are strong indications that other factors than temperature play important roles in regulating the movements of toad larvae.