In the decerebrated and spinalized rabbit, reflexes evoked in the gastrocnemius medialis muscle nerve by electrical stimulation of the sural nerve are suppressed after blockade of NK1-tachykinin receptors. This observation suggests that endogenous tachykinins tonically enhance transmission between sural nerve afferents and gastrocnemius motoneurons. In the present study we have investigated some possible sources of this tachykininergic tone. Electrical stimulation of the sural nerve at 1 Hz, as used in our previous investigation, leads to increased gastrocnemius reflex responses with successive stimuli. We examined reflexes evoked by pairs of sural stimuli separated by intervals of 10-1000 ms, and found that responses to the second stimuli of such pairs were significantly enhanced at intervals from 50 to 500 ms. Treatment with the NK1 receptor antagonist CP-99,994 [(+)-(2S,3S)-3-(2-methoxybenzylamino)-2-phenylpiperidine; 1 mg/kg, i.v.] reduced gastrocnemius reflex responses per se, but did not alter the facilitatory effects of pairing sural stimuli. Subsequent treatment with the glutamate N-methyl-D-aspartate receptor blocker dizocilpine (0.5 mg/kg, i.v.) further reduced reflex responses and abolished paired-pulse facilitation. in a second set of experiments, rabbits were prepared so that reflexes could be studied with minimal surgical preparation of the hindlimb. Reflex responses recorded in this way were enhanced by treatment with CP-99,994 (up to 1 mg/kg, i.v.). Subsequent administration of the opioid receptor antagonist naloxone (1-100 mu g/kg, i.v.) increased reflexes, as seen previously in surgically-prepared animals. These data show that tachykininergic modulation of spinal reflexes in the rabbit results from the nociceptive input arising from surgical preparation of the leg. In contrast, tonic opioidergic inhibition of reflexes is not substantially dependent on such input.