1. The effects of vagal afferent stimulation (VAS) on spinal nociceptive transmission and the spinal pathway(s) mediating VAS-produced effects were examined in pentobarbital sodium- and urethane-anesthetized, paralyzed rats. The 60 units studied responded to mechanical stimuli and noxious heating (50 degrees C) of cutaneous receptive fields confined to the glabrous skin of the toes and footpads. Recording sites were located in laminae I-VI of the L3-L5 spinal segments. 2. VAS facilitated and inhibited neuronal responses to heat. In pentobarbital-anesthetized rats, responses of most (24/44) units were facilitated by low and inhibited by higher intensities of VAS. Responses of some units (15/44) were only inhibited and others (4/44) only facilitated by VAS. Inhibition produced by VAS was intensity-, pulse width-, frequency-, and stimulation duration-dependent. In urethane-anesthetized rats, responses of 6/16 units were initially facilitated, then inhibited as the intensity of VAS was increased; responses of nine units were inhibited by VAS. Quantitative comparisons of recruitment indices, mean thresholds for inhibition and mean intensities to inhibit unit responses to heat to 50% of control revealed no significant differences between the two anesthetic conditions. 3. The effects of VAS on neuronal responses to heat were dissociable from its effect on blood pressure. Regardless of the effect of VAS on unit responses to noxious heat, VAS consistently produced intensity-dependent depressor responses. The latencies to onset of inhibition and facilitation by VAS were determined by a cumulative sum technique and bin-by-bin analysis of peristimulus time histograms. The apparent latencies were 91 +/- 11 (SE) ms for inhibition and 278 +/- 59 ms for facilitation, both of which occurred before changes in blood pressure. Finally, microinjections of lidocaine into the ventrolateral funiculus (VLF) or transections of the dorsolateral funiculus (DLF) of the thoracic spinal cord attenuated VAS-produced effects on neuronal responses, but did not affect VAS-induced depressor responses. 4. The responses of 11 dorsal horn units to graded noxious heating of the skin were studied; the stimulus-response functions (SRF) were linear and monotonic throughout the temperature range examined (42-52 degrees C). VAS at intensities which inhibited unit responses to heat significantly decreased the slope of the SRF. VAS at intensities which facilitated unit responses to heat produced a leftward, parallel shift of the SRF.(ABSTRACT TRUNCATED AT 400 WORDS)