Within the last decade, working memory (WM) has garnered increased interest as a potential core deficit or endophenotype of attention-deficit/hyperactivity disorder (ADHD). The current study is the first meta-analytic review to examine several subject and task moderator variables¿ (e.g., percent female, diagnostic selection procedure, trials per set size, response demands, type of dependent variable, and central executive [CE] demands) effect on between-group phonological (PH) and visuospatial (VS) WM in adults with ADHD, relative to healthy controls. Method: Literature searches were conducted using the PsycINFO, Web of Science, and PubMed databases, and yielded 38 studies of WM in adults with ADHD. Results: Results revealed moderate-magnitude between-group effect sizes (ESs) across both WM domains. In addition, several task-moderating variables explained significant ES variability among PH and VS studies. Conclusions: Collectively, these findings indicate that WM deficits persist into adulthood and suggest that methodological variability may explicate why WM deficits have not been uniformly detected in previous experimental studies. (PsycINFO Database Record (c) 2013 APA, all rights reserved)