Using the conflict drinking test as a model, we studied in rats the effect of the nonselective beta-adrenoceptor blockers pindolol and cyanopindolol which bind to 5-HT1A and 5-HT1B receptors, and of the selective beta 1- and beta 2-adrenoceptor antagonists betaxolol and ICI 118,551, respectively, which have a negligible affinity for 5-HT receptors. Both pindolol (2.0-8.0 mg/kg) and cyanopindolol (0.5-2.0 mg/kg) showed an anticonflict effect, having dose dependently increased the number of punished licks. On the other hand, neither betaxolol nor ICI 118,551--administered separately or in combination--affected the punished responding. The anticonflict effects of pindolol and cyanopindolol were completely abolished by the 5-HT1A receptor and alpha 1-adrenoceptor antagonist 1-(2-methoxyphenyl)-4-[4-(2-phtalimmido)butyl]piperazine (NAN-190), but were not modified by the selective alpha 1(-)-adrenoceptor antagonist prazosin. The effects of pindolol and cyanopindolol were also not modified in animals with lesions of 5-HT neurons, produced by p-chloroamphetamine (PCA). Moreover, it was also found that the anticonflict effects of pindolol and cyanopindolol in PCA-pretreated rats were antagonized by NAN-190 but not prazosin. Our results indicate that the anticonflict effects of pindolol and cyanopindolol depend on their agonist action on postsynaptic 5-HT1A receptors.