The current study is an investigation of the MMPI-2 Fake Bad Scale (FBS) in the detection of incomplete effort in mild head injury (MHI). Using ROC curve analysis, we found that a cutoff score of 21 had a sensitivity of 90% and specificity of 90%, providing an overall correct classificatory rate of 90%. In addition, traditional indices of faking bad on the MMPI-2, the F and F -- K indices, fared relatively poorly by comparison and added no predictive power over the FBS. Finally, multivariate analyses revealed that although the FBS shares a number of items with Hs and Hy scales, the FBS carried the majority of variance in predicting incomplete effort in our MHI sample. Overall, these findings indicate that the FBS has high sensitivity and specificity in identifying incomplete effort in mild head injury. [ABSTRACT FROM AUTHOR]