A computer-assisted version of the delayed matching-to-sample (DMTS) task has been of enormous utility in our non-human primate model for assessment of memory-enhancing agents. To avoid ceiling effects as might be encountered by using fixed delay intervals in monkeys exhibiting varying performance efficiencies, delay intervals are adjusted to provide similar baseline levels of delay-dependent performance. Macaques well trained in the task exhibited a marked age-dependent sensitivity to the effects of the amnestic drug scopolamine. Aged animals also were more affected than their younger counterparts by the presentation of a distractor shortly after receiving the stimulus component of the DMTS task. One limitation of the DMTS task is that under baseline conditions certain aged subjects may perform the task as well or better than some younger animals. To help avoid this situation, we developed a titrating version of the DMTS which was administered similarly to the standard DMTS task. Animals begin the first trial with a 0 s delay interval. Delay intervals after a correct response are incremented by 1 s. Delay intervals after an incorrect match are decreased by 1 s. Rhesus and pigtail macaques who ranged in age from 5-27 years and who were maintained on the standard DMTS for at least one year performed 3-4 consecutive 96 trial sessions. The maximum delay intervals attained by the study group, exhibited a significant correlation with age (p < 0.02). Decrements in task accuracy, and in the number of trials completed/session showed a trend with age (p < 0.08). If the titrating version of the DMTS is sensitive to mnemonic drugs, the task may prove useful for drug comparisons with aging.