Effects of drugs which enhance or reduce GABAergic neurotransmission upon conflict behavior were evaluated with a modified Vogel procedure which was shown to be insensitive to variations in motivation to drink and to the analgesic effects of morphine. In addition, the effects of these drugs on ambulatory activity and motor execution were quantified. For comparison, the benzodiazepines diazepam and chlordiazepoxide were used. Anticonflict actions of diazepam were obtained with a shock current of 0.25 mA but not with 0.05 or 0.5 mA, whereas the proconflict effect of FG7142 was obtained with 0.05 mA but not with higher currents. Diazepam and chlordiazepoxide had anxiolytic effect in a dose similar to that required to reduce ambulatory activity, but below that needed to affect motor execution. At doses high enough to impair motor execution, anticonflict effects were considerable. The GABA-A receptor agonist THIP and the GABA-B receptor agonist baclofen lacked effect on conflict behavior in moderate doses, which reduced ambulatory activity. In doses which produced motor deficiencies these drugs reduced licking both in the conflict test and when tested without shock administration. The effects of the GABA transaminase inhibitors gamma-acetylene GABA and sodium valproate were similar to those of the receptor agonists. The GABA reuptake inhibitor SKF 100330A produced anticonflict effect in a dose below that needed to reduce ambulatory activity, but lacked effect on conflict behavior in higher doses. The GABA antagonist picrotoxin, and the GABA synthesis inhibitors 4-deoxypyridoxine and isoniazide, reduced licking both in the absence and presence of shock, and affected motor functions in the same doses. Bicuculline, at the doses used, had no behavioral effects.(ABSTRACT TRUNCATED AT 250 WORDS)