Background and purpose: Longitudinal studies of mild cognitive impairment (MCI) report that a sizeable proportion of MCI cases revert to normal levels of functioning over time. The rate of recovery from MCI indicates that existing MCI diagnostic criteria result in an unacceptably high rate of false positive diagnoses and lack adequate sensitivity and specificity.

Methods: The aim of the present study was to identify a set of neuropsychological measures able to differentiate between true positive cases of MCI from those who were unimpaired at 11 months’ follow-up.

Results: A discriminant function analysis identified that a combination of measures of complex sustained attention, semantic memory, working memory, episodic memory and selective attention correctly classified outcome in more than 80% of cases. The rate of false positive diagnoses (5.93%) was considerably lower than is evident in previously published MCI studies.

Conclusions: The results of the present study indicate that the rate of false positive MCI diagnoses can be significantly reduced through the use of sensitive and specific neuropsychological measures of memory and non-memory functions.