1. Electromyographic recordings were made from the biceps femoris muscle through a pair of noninsulated platinum/iridium needle electrodes in male Sprague-Dawley rats artificially ventilated and anesthetized with 0.8% halothane in a N2O-O2 mixture (2/3:1/3). The animals' ventilation, heart rates, and body temperatures were continuously monitored. Electrical stimuli (2-ms duration, 0.16 Hz) were delivered to the sural nerve territory through a pair of noninsulated platinum/iridium needle electrode inserted subcutaneously in the medial aspect of the 4th and the lateral part of the 5th toe. Such stimulation elicited a two-component reflex response in the ipsilateral biceps femoris muscle: The first had a short latency (17.5 +/- 2.3 ms), short duration (20.7 +/- 2.6 ms), and low threshold (1.5 +/- 0.6 mA), whereas the second had a longer latency (162.4 +/- 5.1 ms), longer duration (202.3 +/- 6.2 ms), and higher threshold (5.7 +/- 0.5 mA). 2. Lidocaine (0.02–0.1%; 0.1 ml), but not saline, injected subcutaneously over the proximal part of the sural nerve, produced a selective depression of the late component of the reflex response, whereas the first component remained unchanged. The conduction velocity of the afferent fibers was estimated from the stimulation needles in the sural nerve territory to the nerve's projection in the lumbar spinal cord: it was concluded that the second, late component of the reflex response was due to afferent signals transmitted via unmyelinated C-fibers, whereas the first component was related to activation of fine myelinated fibers (A delta group). 3. Electrical stimulation of the sural nerve was still able to elicit the two-component reflex responses in the ipsilateral biceps femoris muscle of chronic spinal rats, indicating that these responses were genuine reflex responses, transmitted completely through a spinal circuit. 4. The C-fiber reflex was recorded when the duration and frequency of the stimuli applied to the sural nerve varied within the 0.5- to 4-ms and 0.02- to 1-Hz ranges, respectively. It was concluded that a single 2-ms duration shock at an intensity of 1.2 times the C-fiber reflex threshold, delivered every 6 s (0.16 Hz), constituted an acceptable and optimal protocol for experiments in which the C-fiber reflex was studied as a function of time. These parameters were used throughout the subsequent experiments.(ABSTRACT TRUNCATED AT 400 WORDS)