The present experiment assessed dual-task performance in 20 young (mea
n age 21) and 20 old (mean age 72) adults. Ss first received extensive
single-task practice on consistent and varied search tasks. Next, the
y received dual-task practice in 2 conditions: (a) varied visual searc
h plus varied memory search and (b) consistent visual search plus vari
ed memory search. In the varied-varied condition, young and old adults
showed similar dual-task decrements. These results, along with the cu
rrent data in the literature, suggest that practice may play an import
ant role in determining age-related dual-task differences (or lack the
reof). In the consistent-varied condition, young adults attained singl
e-task performance levels, because they had automatized the consistent
task. Old adults were unable to match their single-task performance l
evels, indicating that they were unable to automatize the consistent t
ask.