Performance of three rhesus monkeys on a test of one-trial stimulus-reward association, in which recall intervals ranged from 0.5 to 6.5 min, was evaluated during nondrug-control conditions and following administration of the muscarinic-receptor blocker scopolamine. During control sessions, performance averaged 78% correct responses. Following administration of 10.0 and 17.8 micrograms/kg of scopolamine, performance fell significantly, to 69% and 63% correct responses, respectively. This dose-dependent impairment in recall was similar to the impairment we reported previously in recognition. Although the results thus failed to support a suggestion derived from behavioral electrophysiological findings that stimulus-reward association might be more vulnerable to scopolamine than stimulus recognition, they provide additional evidence for a cholinergic contribution to cognitive memory.