Younger and older adults were administered a computerized version of the Corsi Block visuospatial working memory (VSWM) span task at either their peak or off-peak time of day and in either a high-interference (ascending order of administration, starting with short lists first) or low-interference (descending order, starting with longest lists first) format. Young adults' span scores were highest in the ascending format. By contrast, older adults performed better in the low-interference format, replicating findings with verbal memory span studies. Although both age groups benefited from being tested at their peak time, the advantage was far greater for older adults, but only in the low-interference format; their scores on the high-interference format were not helped by peak-time testing. These findings are consistent with the suggestion that young adults' performance on span tasks is influenced by practice and strategies, but the performance of older adults is heavily influenced by interference-which is best controlled at peak times of day. Our findings suggest that both time of testing and interference play critical roles in determining age differences in VSWM span, and both a reduction in interference and peak-time testing may be necessary to optimize older adults' performance and to maximize the reduction in age differences.