The effects of sudden cooling of the spinal cord were studied in three species of amphibians--a cold-sensitive tropical toad (Bufo marinus), a cold-resistant, aquatic, hibernating frog (Rana pipiens, northern leopard frog) and a freeze-tolerant frog (Rana sylvatica, wood frog). Ventral root (motoneuron) potentials were recorded from isolated, hemisected spinal cords of each species mounted in a sucrose-gap recording apparatus and superfused with HCO3(-)-buffered Ringer's solution at room temperature (21 degrees C). In the toad, sudden cooling to 6-8 degrees C produced large, sustained motoneuron depolarizations that returned slowly to baseline levels and were accompanied by extensive paroxysmal activity. Larger, but shorter-lasting, motoneuron depolarizations associated with only a limited amount of paroxysmal activity were generated by rapid cooling of the leopard frog spinal cord. Small, brief motoneuron depolarizations followed by a hyperpolarization, or hyperpolarizations not preceded by depolarizations, were seen in cooled wood frog spinal cords. The wood frog displayed a large amount of spontaneous motoneuron activity, but little paroxysmal activity in response to sudden cooling. Following prolonged cooling, rewarming the spinal cords of all three species resulted in motoneuron hyperpolarizations that slowly decayed towards the baseline value. The amplitude of the rewarming-induced response was larger and longer in toad motoneurons than in leopard frog and wood frog motoneurons. At room temperature, a single supramaximal dorsal root stimulus evoked a depolarizing ventral root potential in toad and leopard frog motoneurons that was decreased in amplitude and prolonged when the spinal cords were cooled to 8 degrees C or below. In contrast, at room temperature, the ventral root reflex in the wood frog was followed by a distinct hyperpolarization. Cooling the wood frog spinal cord only slightly reduced the amplitude of the ventral root potential. In contrast, the evoked hyperpolarization was blocked by sudden cooling and also by the addition of dihydro-ouabain to the Ringer's solution. The motoneuron hyperpolarizations induced by sudden cooling in the wood frog were converted to depolarizations when Cl- in the superfusate was replaced with isethionate. The depolarizations elicited by sudden cooling were reduced by the addition of kynurenate in all three species. A dose-response curve generated by short applications of L-glutamate demonstrated that wood frog motoneurons were less sensitive than leopard frog motoneurons to L-glutamate. In summary, three species of amphibians, differing in their adaptations to the temperature of their environments, vary in their responses to sudden reductions in temperature. The relationship of these responses to their environmental adaptations remains to be determined.