In freshly isolated spinal dorsal horn (DH) neurons (laminae I–IV) of the young rat, the effects of dynorphin A 1–17, U-50,488H and U-69,593 on inward currents induced by α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and kainate (KA) were studied under whole-cell voltage-clamp conditions. When the cells were clamped to a holding potential of −60 mV, co-application of dynorphin A 1–17 (10 −6 M) and AMPA (2 × 10 −5 M) reversibly decreased the peak amplitude of the initial transient component of the AMPA-induced current in 72% of the examined cells. In addition, dynorphin (10 μM) in perforated patch-recordings consistently produced a decrease in the steady-state component of the AMPA response. The depressant effect was concentration-dependent (IC 50 = 86 nM) and reversible. The dynorphin A 1–17-induced depression of the AMPA response was associated with slowing of the response kinetics, including both a 10–90% rise-time and time constant of decay. The AMPA-induced currents were modulated by dynorphin not only during the co-administration but also after the removal of the peptide. Dynorphin increased the initial peak AMPA current in 42% of the examined cells. Similar as with dynorphin A 1–17, the peak amplitude of the AMPA-induced current was reversibly suppressed in the presence of 1 μM U-50,488H and U-69,593 in 75% and 86% of the examined cells, respectively. Naloxone and the κ 1-selective antagonist norbinaltorphimine (nor-BNI) blocked the initial depressant but not late excitatory effects of dynorphin A 1–17 and U-50,488H. This antagonistic effect of naloxone and norbinaltorphimine suggests that the depressant effect of dynorphin A 1–17 on the AMPA-activated conductance is a true opioid, probably κ 1-opioid receptor-mediated event. In contrast, the dynorphin-induced late potentiation of AMPA/KA responses appears to be a non-opioid effect since it was not inhibited by nor-BNI, CTAP and naltrindole, the selective κ-, μ- and δ-opioid receptor blocking agents, respectively. Pretreatment of DH neurons with pertusis toxin blocked the depressant action of dynorphin A 1–17, indicating that a G i- or G 0-type G protein was required for this effect on AMPA-activated currents. Intracellular dialysis with a highly specific peptide inhibitor (peptide 6–22) of the cAMP-activated protein kinase (PKA), and with Rp-cAMPS, prevented the depressant efect of dynorphin A 1–17. In addition, staurosporine, a nonselective kinase inhibitor, blocked the dynorphin depression of the AMPA response. These results suggest the possibility that dynorphin, acting through a decrease in intracellular cyclic AMP levels, can reduce the responses of DH neurons to exogenous AMPA. Besides modulating the AMPA response of DH cells, co-application of 1 μM dynorphin and KA (2–5 × 10 −5 M) decreased the magnitude of the KA-induced current in 50% of the cells tested and increased it upon the removal of the peptide. Nor-BNI and Rp-cAMPs prevented the depression of KA responses, whereas the late potentiation was not modified. Our results suggest that dynorphin A 1–17, U-50,488H and U-69,593 modulate the AMPA/KA receptors signaling function in a subset of the rat spinal DH neurons. Possible mechanisms of action are discussed in relation to dynorphin-depressant regulation of excitatory amino acid-mediated primary afferent neurotransmission, including nociception.