This study investigated the utility of four WAIS-IV Digit Span (DS) indices (traditional Reliable Digit Span [RDS], RDS-Working Memory [RDS-WM], RDS-Revised [RDS-R], and DS Age-Corrected Scaled Score [ACSS]) as embedded performance validity tests (PVTs) among a sample of 342 consecutive adults referred for neuropsychological evaluation of ADHD. All DS indices had acceptable classification accuracy (areas under the curve: 0.73-0.76) for detecting invalid performance with optimal cut-scores of RDS [less than or equal to]7 (35% sensitivity/93% specificity), RDS-WM [less than or equal to]7 (56% sensitivity/86% specificity), RDS-R [less than or equal to]12 (48% sensitivity/85% specificity), and ACSS [less than or equal to]7 (46% sensitivity/87% specificity). Although all indices were able to detect invalid performance, DS indices incorporating the more complex working memory trials of the task yielded the best accuracy for identification of invalid test performance among adults referred for ADHD evaluation.