This article reviews automated test procedures that can be used to stu
dy memory for discrete/trial-unique events in rodents and other animal
s. A distinction is made between delayed response tasks, i.e. those ta
sks for which all information necessary to determine the correct respo
nse is available prior to the retention interval, and delayed comparis
on tasks, i.e. tasks in which the subject must compare stimuli present
ed prior to and after the retention interval in order to determine the
correct response. Delayed comparison procedures potentially provide a
purer estimate of the subject's memory capacities, but have also been
more difficult for rodents to acquire. Suggestions are offered for po
tential directions for improved delayed comparison tasks.