New methodologies to assess analgesic response in humans are needed to better integrate preclinical and clinical data. In the present study we examined the test-retest stability of an innovative radiant heat methodology compared with an electrical stimulation methodology. For the radiant heat task, a modified rodent tail flick apparatus was used, The latency for fmger withdrawal was recorded. For the electrical stimulation tasks, subjects placed two fingers on two electrodes from which they received a brief series of increasingly intense electrical stimulations. Maximumstimulus intensity (in milliamps) delivered was recorded. On each of 4 test days, the subjects received five test trials with a lO-min interval between trials. Allthe subjects were tested twice on each apparatus in a counterbalanced design. Finger withdrawal latencies for the radiant heat task did not differ significantly across test trials or test days. Finger withdrawal scores for electrical stimulation increased significantly across test trials as well as test days. These data show that the radiant heat method generates consistent latencies across trials and days, whereas shock produces trends over time. The radiant heat task, which is convenient to operate and inexpensive to build, appears promising as a reliable test of pain threshold in humans.