The potency of 7 excitatory amino acid antagonists had been measured at kainate receptors on primary afferent C fibres using isolated dorsal roots from immature rats. Two of the compounds were tested as antagonists of excitant amino acids at motoneurones using isolated hemisected spinal cord preparations. Mean dose-ratios for antagonism of kainate at C fibres, produced by 1 mM antagonist in at least four preparations, ranged from 2.1 +/- 0.5 (mean +/- s.e.mean) with 2-amino-5-phosphonopentanoate (AP5) to 17.6 +/- 2.4 with 1-(p-bromobenzoyl)-piperazine-2, 3-dicarboxylate (BBpzD). The rank order of potency of antagonists at C fibres was similar to that obtained previously for antagonism of kainate at motoneurones. The potency of kynurenate as an antagonist of kainate at C fibres (apparent Kd 70 +/- 4.3 microM (mean +/- s.e.mean), n = 12) was significantly different (P less than 0.005, Wilcoxon rank sum) from its potency at motoneurones (apparent Kd 164 +/- 14 microM, n = 13). Kynurenate also was significantly (P less than 0.025 Wilcoxon rank sum) more potent at antagonizing kainate- than quisqualate (apparent Kd 258 +/- 28 microM, n = 12)-induced depolarization of motoneurones. Kynurenate and BBpzD (0.25, 1.0 and 4.0 mM) were compared as antagonists of N-methyl-D-aspartate (NMDA) at motoneurones and the slope of the Gaddum-Schild plot for kynurenate was markedly greater than 1 (2.01 +/- 0.22, 95% confidence limits). A greater than additive antagonism of NMDA-induced depolarizations was produced by combinations of kynurenate with, either AP5, or magnesium ions. 5. The results suggest that the kainate receptor on C fibres may be different from the kainate receptor on motoneurones and that antagonism of NMDA-induced depolarization by kynurenate may not be entirely competitive.