Lesions of the amygdala impair performance on a conditioned place preference (CPP) but not a spatial radial maze task. The role of cholinergic receptors within the amygdala in performance of these tasks was evaluated using intra-amygdala injections of the muscarinic receptor antagonist, scopolamine. Food deprived rats were trained on a CPP task, which consisted of four training trials on two arms of a radial eight-arm maze. One arm was consistently paired with a large amount of food (14 g) while the other arm was never baited. Prior to the fourth trial, rats received bilateral intra-amygdala infusions of the muscarinic receptor antagonist, scopolamine (SCOP; 5 microg/0.5 microl) or vehicle. On a retention test 24 h later, unoperated and vehicle-infused rats, but not SCOP-treated rats, spent significantly more time in the paired arm than chance (50%). Therefore, the scopolamine treatment appeared to block learning and/or memory on trial 4. The same rats were then trained on a radial maze task on the same apparatus, in which rats had access to all eight arms but only four were baited with food (1 pellet). Rats were trained until they reached criterion and then infusions were given prior to testing. SCOP treatment did not affect performance on the radial maze task. Thus, intact cholinergic mechanisms in the amygdala are necessary for learning or memory on a CPP task with a high reward component but not performance on a spatial radial maze task with a lower reward component.