Cebus monkeys of 3 different age groups were trained to perform an automated behavioral task (delayed response), intended to measure recent memory ability. In in initial study, the aged monkeys (18 years and older) exhibit prprogressively greater performance impairments (relative to young monkeys) as they were required to remember the location of a visual stimulus for increasingly longer durations (0 to 20 sec). This deficits replicated previously published results from aged Rhesus monkeys and appeared similar to the primary memory deficits reported in elderly humans and demented patients. In subsequent studies, the effects of three different cholinomimetics were evaluated for their ability to improve the aged monkey's performance on this task. Each monkey was tested under several acute doses of the cholinergic precursor, choline, the anticholinesterase, physostigmine, and the cholinergic muscarinic receptor agonist, arecoline. The results revealed clear differences in the ability of these drugs to improve performance on this task. Choline exerted no apparent effects in the aged monkeys at any dose tested. Physostigmine clearly enhanced performance in certain aged monkeys, but the optimal dose varied dramatically between subjects, replicating previously published results with aged Rhesus monkeys and humans. Arecoline produced clear improvement within a restricted dose range, with little variation in optimal dose between subjects. In addition to demonstrating differences in the effects of different cholinomimetics on memory performance in aged primates, these data also suggest a possible rationale for future investigations. Assuming that each of these drugs primarily affected cholinergic function in the manner conventionally attributed, these data suggest that, within the cholinergic system, the more directly one stimulates the receptor, the more one might expect robust and consistent effects on memory performance in aged subjects.