1. We investigated the role of nitric oxide (NO) in modulating spinal synaptic responses evoked by electrical and noxious sensory stimuli in the neonatal rat spinal cord in vitro. 2. Potentials were recorded extracellularly from a ventral root (L3-L5) of the isolated spinal cord preparation or spinal cord-saphenous nerve-skin preparation of 0- to 2-day-old rats. Spinal reflexes were elicited by electrical stimulation of the ipsilateral dorsal root or by noxious skin stimulation. 3. In the spinal cord preparation, single shock stimulation of a dorsal root at C-fibre strength induced mono-synaptic reflex followed by a slow depolarizing response lasting about 30 s (slow ventral root potential; slow VRP) in the ipsilateral ventral root of the same segment. Bath-application of NO gas-containing medium (10(-4)- 10(-2) dilution of saturated medium) and NO donors, 1-hydroxy-2-oxo-3-(N-ethyl-2-aminoethyl)-3-ethyl-1-triazene (NOC12, 3-300 microM), S-nitroso-N-acetyl-D,L-penicillamine (SNAP, 3-300 microM) and S-nitroso-L-glutathione (GSNO, 3-300 microM), produced an inhibition of the slow VRP and a depolarization of ventral roots. Another NO donor, 3-morpholinosydononimine (SIN-1, 30-300 microM), also depressed the slow VRP but did not depolarize ventral roots. These agents did not affect the mono-synaptic reflex. 4. In the spinal cord-saphenous nerve-skin preparation, application of capsaicin (0.1-0.2 microM) to skin evoked a slow depolarizing response of the L3 ventral root. This slow VRP was depressed by NOC12 (10-300 microM) and SIN-1 (100-300 microM). When the concentration of NOC12 was increased to 1 mM, spontaneous synaptic activities were augmented and the depressant effect of NOC12 on the slow VRP became less pronounced. 5. A NO-scavenger, 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide( carboxy- PTIO, 100-300 microM) prevented the depressant effect on the dorsal root-evoked slow VRP and ventral root depolarizing effects of NO donors. Carboxy-PTIO increased spontaneous synaptic activities and markedly potentiated the slow VRP. A NO synthase (NOS) inhibitor, N omega-nitro-L-arginine methyl ester (L-NAME, 0.03-1 microM), but not D-NAME (0.03-1 microM), also markedly potentiated the slow VRP and this effect was reversed by L-arginine (300 microM). 6. 8-Bromo-cyclic guanosine 3': 5'-monophosphate (8-Br-cyclic GMP, 100-300 microM) produced both an inhibition of the slow VRP and a depolarization of ventral roots. A cyclic GMP-dependent protein kinase inhibitor, KT5823 (0.3 microM), partly inhibited the depressant effects of NO donors and 8-Br-cyclic GMP on the dorsal root-evoked slow VRP. In contrast, KT5823 did not inhibit the depolarizing effects of NO donors. 7. Perfusion of the spinal cord with medium containing tetrodotoxin (0.3 microM) and/or low Ca2+ (0.1 mM)-high Mg2+ (10 mM) markedly potentiated the depolarizing effect of NO donors. The SNAP-evoked depolarization in the tetrodotoxin-containing low Ca(2+)-high Mg2+ medium was significantly inhibited by excitatory amino acid receptor antagonists D-(-)-2-amino-5-phosphonovaleric acid (30 microM) and 6-cyano-7-nitroquinoxaline-2,3-dione (10 microM). 8. The present study suggests that inhibitory and excitatory mechanisms meditated by the NO-cyclic GMP cascade are involved in the primary afferent fibre-evoked nociceptive transmission in the neonatal rat spinal cord. The inhibitory mechanism, but not the excitatory mechanism, appears to be partly mediated by cyclic GMP-dependent protein kinase. It is also suggested that Ca(2+)-independent release of excitatory amino acid neurotransmitters contributes to the depolarizing response to NO of ventral roots.