Validation studies of a 2-alternative forced-choice test used to detect faked cognitive impairment are reported. In Study 1, 177 college students were given substantial financial incentives to fake believable impairment. Rate of detection was compromised by financial incentives, but the test demonstrated superior specificity and sensitivity relative to other measures of response bias. Study 2 and Study 3 included neuropsychology (n = 134) and forensic (n = 18) evaluees, who were administered several tests of response bias. One decision rule for the forced-choice test demonstrated greater sensitivity than the other measures of response bias and displayed moderately good agreement with clinician ratings of individual testing response style. Although we originally intended to identify malingering, we conclude that psychometric tasks can only detect biased responding and that the determination of malingering must be based on overall clinical evaluation.