During personality inventory construction, researchers often exclude items that correlate strongly with a social desirability measure. The validity of this practice, however, may vary depending on the choice of social desirability measure and this concern has not been thoroughly evaluated. To redress this methodological gap, the current research examines the validity of using the Over-Claiming Technique (OCT; Paulhus, Harms, Bruce, & Lysy, 2003) for probing item social desirability. Results indicate that item desirability values gauged with the OCT showed weak evidence for validity; they correlated negatively, rather than positively, with judges' direct ratings of personality item desirability. Therefore, a recommendation is developed against the usage of the OCT in gauging item desirability. Possible reasons for the weak validity evidence and future research directions are discussed.