The participants with normal memory were 40 college students at West Virginia University. They were randomly assigned to either experimental or control conditions and were compared with 20 patients with memory impairment, as shown by scores on either of two standard instruments, who were given the procedure under instructions to perform at their best capacity. The tests were conducted in a sound-attenuated laboratory. The memory test consisted of reading and recalling 21-item word lists. Results showed that the students who had been instructed to falsify memory impairment showed different patterns of responding and could be clearly distinguished from students performing their best and from persons with actual memory impairment. Thus, a 100-percent rate of correct classifications occurred for normal and memory-impaired persons, compared to 65 percent for the malingerers. Tables and 18 references (Author abstract modified)