Brain serotonergic, noradrenergic and GABAergic mechanisms are all involved in the regulation of conflict behaviour, and the GABAA/benzodiazepine receptor complex may play the most central role in this context. Since facilitation of GABAergic inhibitory transmission produces anticonflict effects, it has been suggested that antagonism of excitatory inputs may serve the same cause, and, indeed, blockade of excitatory neurotransmission mediated via N-methyl-D-aspartate (NMDA), receptors, produces anticonflict effects. In the present study, using a modified Vogel's rat conflict model, we have investigated whether the anticonflict effect of the non-competitive NMDA antagonist MK-801 can be linked to NMDA receptor blockade, and if stimulation of these receptors instead produces proconflict effects. The tentative involvement of noradrenergic, serotonergic or GABAergic effects in the MK-801-induced anticonflict effect was also studied. MK-801 produced a dose-dependent and specific anticonflict effect (maximal effect after 0.05 mg/kg, intraperitoneally, -90 min.). This anticonflict action was completely counteracted by NMDA in a dose (0.125 microgram, intracerebroventricularly) not affecting behaviour per se. The highest dose tested of NMDA alone (0.5 microgram) tended to produce a proconflict effect, but this action may be unspecific due to concomitant drug-induced motor-inhibition. Neither bicuculline and picrotoxin, antagonists at the GABAA/benzodiazepine receptor complex, nor the adrenoceptor antagonists propranolol and prazosin significantly altered the MK-801-induced anticonflict effect, whereas L-5-HTP (50 mg/kg, intraperitoneally, after inhibition of peripheral decarboxylation with benzerazide) completely abolished the anticonflict effect of MK-801.(ABSTRACT TRUNCATED AT 250 WORDS)