In the rat the effects of clonidine and xylazine were compared on a measure of analgesia (tail-withdrawal from hot water) and on operant responding maintained by a fixed-ratio 20-response schedule of food presentation. Clonidine (2--8 mg/kg) produced dose-dependent increase in tail-withdrawal latency and was approximately twice as potent as xylazine (8--16 mg/kg) in this test. This analgesic effect of 8 mg/kg of clonidine was antagonized by phenoxybenzamine (10 mg/kg), and high doses of yohimbine (5--10 mg/kg), whereas the effect of 2 mg/kg of clonidine was potentiated by a nonanalgesic dose of nisoxetine (10 mg/kg). Clonidine (0.0063--0.2 mg/kg) and xylazine (0.25--8 mg/kg) produced a dose-dependent suppression of fixed-ratio responding. The potency of clonidine in this task was approximately 40 times greater than that of xylazine. The suppression of responding produced by both drugs was antagonized by low doses of yohimbine (0.5--2 kg/mg), which was maximally effective at a dose of 1 mg/kg. The data suggest that when using this type of assay for analgesia, the antinociceptive effects of clonidine which are measured are a result of stimulation of alpha-adrenergic receptors or noradrenergic neurons or by adrenergic receptors which inhibit noradrenergic neurons.