The effect of intracerebroventricular (i.c.v.) injection of dynorphin A-(1-13), an endogenous kappa-opioid receptor agonist, on memory process was examined in mice using elevated plus-maze test. Dynorphin A-(1-13) (1, 3 and 10 mug) did not influence the transfer letency on retention test. In contrast, dynorphin A-(1-13) (10 mug) significantly improved the memory impairment induced by socopolamine (1 mg/kg s.c.). The effect of dynorphin A-(1-13) (10 mug) on the scopolamine-induced amnesia was almost completely reversed by pretreatment with nor-binaltorphimine (4 mug, i.c.v.), a kappa-selective opioid antagonist. In addition, RU 24213 (1 and 3 mg/kg), a dopamine D2-selective agonist, almost completely reversed the improving effect of dynorphin A-(1-13) (10 mug) on the scopolamine-induced amnesia. However, SKF 38393 (3, 10 and 30 mg/kg), a dopamine D1-selective agonist, failed to change the transfer latency on retention test in scopolamine-treated mice. These findings suggest that the improving effect of dynorphin A-(1-13) on the scopolamine-induced amnesia may be based upon the inhibition of dopaminergic activity through the mediation of kappa-opioid receptors.