Neuropsychological baseline testing is commonplace in the assessment of concussion; however, claims of "sandbagging" the baseline have led neuropsychologists to ask to what extent athletes can perform intentionally poorly on baseline testing without reaching threshold on the test validity indicators. Seventy-five undergraduate athletes were re-administered the ImPACT neurocognitive battery, which they had previously taken to establish baseline functioning, but were instructed to perform more poorly than their baseline without reaching threshold on the test validity indicators. Eight participants were able to successfully fake significantly lower scores without detection by validity indicators. Concussion history was not related to performance. Successful fakers did not perform significantly worse on the Reaction Time Composite and Three Letters Total Letters Correct, questioning the utility of these measures for detecting "sandbagging." Successful fakers reported using less purposeful faking strategies which naturally facilitated errors. The data suggest that "sandbagging" the baseline, even under conditions involving motivation, instruction, and experience with the test, is difficult to accomplish without being detected.