The detection of dissimulated cognitive deficits is an important issue in many clinical assessment situations. Specific procedures designed to aid in this detection can be useful additions to standard assessment protocols. The present study investigates the potential utility of a modification of the forced-choice, symptom validity paradigm to the assessment of memory deficits. Previous research indicated that the procedure could discriminate between college students instructed to malinger memory impairment and college students instructed to perform at their best ability. The present study administered a forced-choice memory assessment instrument to 60 psychiatric patients, 60 normal community volunteers, and 60 neuropsychological patients. Thirty psychiatric patients and 30 community volunteers were given instructions to malinger. Discriminant function analyses yielded an overall correct group classification rate of 90.6%, with 5% of the nonmalingering subjects and 18.3% of the experimental malingerers incorrectly classified.