1. To quantitatively investigate a nocifensive behavioral response, we developed a method to measure the magnitude of the rat's tail flick reflex and its modulation. A radial array of force transducers measured forces of tail flicks (in rostral, horizontal, and vertical planes) elicited by graded noxious radiant thermal stimulation of the conscious rat's tail, from which the overall movement vector was calculated. 2. The rostrally directed component of tail flicks was always larger than dorsal or horizontal components; the latter was usually in a preferred (left or right) direction regardless of which side of the tail was heated. Tail flick force vectors increased from 40 to 46-52 degrees C and then leveled off. Stimulus-response functions were reproducible within and across rats and were fitted by second-order polynomial functions, whose correlation coefficients were similar when the left or right side of the tail was stimulated in separate sessions (r2 = 0.408 and 0.451, respectively). The inverse latency of tail flicks also increased with temperature in a manner fitted by a second-order polynomial (r2 = 0.707, 0.553 for left and right side, respectively). 3. Systemic administration of morphine (1 or 2 mg/kg ip) usually suppressed tail flicks in an all-or-none manner; i.e., flicks at all stimulus temperatures were either totally abolished (n = 7) or unaffected (n = 5) after morphine. In three rats, 1 mg/kg morphine suppressed tail flick magnitude subtotally, reducing the slope of the linear portion of the stimulus-response function. Morphine effects were reversed by the opiate antagonist naloxone. 4. Tail flick magnitude decreased over repeated trials of 44 degrees C heat stimuli delivered to one tail site, recovered after a 15-min rest period, and decremented more quickly with subsequent stimulus repetition. The decrement was less at long (2 or 4 min) than at short (1 min) interstimulus intervals, and high (50 degrees C) than at low (44 degrees C) stimulus intensities. The reflex decrement transferred to a nearby stimulus site in some rats, and was "dishabituated" after a noxious tail pinch. These observations are consistent with habituation of the tail flick reflex. 5. This method, therefore, provides a quantitative and reproducible measure of tail flick reflex magnitude that is sensitive to morphine. The underlying neural circuitry of the tail flick reflex is discussed in relation to limb withdrawal reflexes.