The effects of cholinomimetic and non-cholinomimetic agents on spatial memory using a novel task in the 12-arm radial maze were investigated. The task was designed to reduce the tendency to use non-spatial strategies. Animals were repeatedly trained to retrieve food rewards from three arms, until a criterion level of performance was reached. Scopolamine (0.03 and 0.1 mg/kg SC), but not N-methylscopolamine (0.1 mg/kg SC) disrupted performance of this task. Physostigmine (0.3 mg/kg SC) and pilocarpine (30 mg/kg SC) completely reversed the deficit of performance produced by scopolamine. Furthermore, the ACE inhibitor Hoe 288 (10 nmol ICV) and the angiotensin AT1 receptor antagonist losartan (10 mg/kg SC) also significantly attenuated the scopolamine-induced deficit. These results show that this novel task in the radial maze is sensitive to the disruptive effects of scopolamine and can identify cognitive enhancing effects of both cholinomimetic and non-cholinomimetic drugs. Thus, this maze task provides a useful model for the evaluation of novel cognitive enhancing agents.