During the last 10 years, computerized dynamic posturography has yielded various patterns of sway on the sensory organization test and the motor control test that have been associated with a variety of organic balance disorders. Some aspects of performance during computerized dynamic posturography, however, are under conscious control. Voluntary movements not indicative of physiologic response to balance system stimulation can also affect computerized dynamic posturography results. Quantification of nonorganic or 'aphysiologic' response patterns in normal subjects, patients, and suspected malingerers is crucial to justify use of computerized dynamic posturography for identification of physiologically inconsistent results. For this purpose the computerized dynamic posturography records of 122 normal subjects, 347 patients with known or suspected balance disorders, and 72 subjects instructed to feign a balance disturbance were critically evaluated by use of seven measurement criteria, which were postulated as indicating aphysiologic sway. Each criterion was scored with a standard calculation of the raw data in a random, blinded fashion. The results of this multicenter study show that three of the seven criteria are significantly different in the suspected 'malingerer' group when compared with either the normal or patient group. The relative strength of each criterion in discerning organic from nonorganic sway provides the examiner with a measure of reliability during platform posture testing. This study demonstrates that computerized dynamic posturography can accurately identify and document nonorganic sway patterns during routine assessment of posture control.