Objectives: To provide a systematic review of age-related differences in n-back performance. Method: Meta-analytic data aggregation. Results: Access for items stored within the focus of attention (0-back and 1-back) was very fast and quasi-perfect; when items are held outside the focus (n > 1), an additional cost was accrued in both accuracy and response time. Age-related differences in accuracy conformed to this bifurcation. Longer lists led to larger costs when going from 1-back to 2-back in older adults. For 1-back accuracy, studies that used visual (as opposed to verbal) stimuli, were experimenter-paced, and used shorter list lengths led to larger age-related differences; for 2-back accuracy, a larger difference in chronological age, visual stimuli, and a higher target proportion led to larger age-related differences. For 1-back response times, age-related differences were larger for studies that had a larger chronological age difference, used experimenter pacing, did not contain lures, and used shorter list lengths; for 2-back response times, age-related differences were larger for studies with larger chronological age differences, visual presentation, experimenter pacing, and a higher target proportion. Discussion: The results suggest a specific age-related deficit associated with focus switching within working memory. Evidence for specific executive-control-related explanations, in contrast, is mixed.