Working memory capacity has been shown to be related to some IAT D measures and has also been suggested to be involved in faking processes. However, research has yet to investigate whether working memory load (WML) impacts all Dmeasures in the same way and how WML impacts faked IAT D measures. Thus, the current study investigated the impact of WML on nonfaked and faked IAT D measures by randomly assigning 48 participants to one of two experimental groups. In Session 1, both groups took the IAT under standard instructions (baseline). In Session 2, both groups took the IAT while simultaneously working on the random number generation (RNG) task to stress their WM (baseline under WML). In Session 3, both groups were given faking strategies and were asked to fake either high or low scores on the IAT (faking ). In Session 4, both groups were asked to fake again while simultaneously working on the RNG task (faking under WML). The results revealed similar impacts of WML on D measures under nonfaking conditions. Under faking conditions, WML somewhat attenuated the effect of the faking of high scores (i.e., the more difficult faking condition) but did not impact the faking of low scores. Moreover, the results imply that the faking strategies of slowing down versus accelerating are associated with different aspects of WM, which provides more detailed insight into the faking process process.