The effects of different manipulations of brain serotonergic (5-HT) systems were investigated in Montgomery's conflict test, an animal anxiety model based on the animal's inborn urge to explore a new environment and its simultaneous fear of elevated, open spaces. The putative 5-HT1A receptor agonists buspirone, gepirone, ipsapirone and 8-OH-DPAT all produced anxiolytic-like effects in narrow low dose-ranges, while in higher doses the behavior returned towards that seen in controls and, after the highest doses of buspirone and gepirone, was suppressed below that of controls. The 5-HT precursor L-5-HTP produced a biphasic dose-response curve. In a single low dose an anxiolytic-like action was obtained, while in the highest dose a clear-cut anxiogenic-like action was observed. The 5-HT depleting agents parachlorophenylalanine (PCPA) and 5,7-dihydroxytryptamine (5,7-DHT) both produced anxiolytic-like effects. Thus, anxiolytic-like and anxiogenic-like effects after various manipulations of brain 5-HT neurotransmission can be obtained also in an animal anxiety model involving neither consummatory behavior nor punishment. The anxiolytic-like effects after all these treatments are suggested to be due to decreased brain 5-HT neurotransmission, while the anxiogenic-like effect obtained after the highest dose of L-5-HTP is suggested to derive from increased 5-HT neurotransmission. Furthermore, using a modified Vogel's conflict model it was investigated whether the anxiolytic-like effects obtained after PCPA and 5,7-DHT involve the GABAA/benzodiazepine (BDZ) chloride ionophore receptor complex. Both the BDZ receptor antagonist flumazenil and the GABAA receptor antagonist bicuculline counteracted the PCPA induced anticonflict effect in doses which did not affect the behavior per se.(ABSTRACT TRUNCATED AT 250 WORDS)