Previous studies of mild cognitive impairment (MCI) have been criticised for using the same battery of neuropsychological tests during classification and longitudinal followup. The key concern is that there is a potential circularity when the same tests are used to identify MCI and then subsequently monitor change in function over time. The aim of the present study was to examine the evidence of this potential circularity problem. The present study assessed the memory function of 72 MCI participants and 50 healthy controls using an alternate battery of visual and verbal episodic memory tests 9 months following initial comprehensive screening assessment and MCI classification. Individuals who were classified as multiple-domain amnestic MCI (a-MCI+) at screening show a significantly reduced performance in visual and verbal memory function at followup using a completely different battery of valid and reliable tests. Consistent with their initial classification, those identified as nonamnestic MCI (na-MCI) or control at screening demonstrated the highest performance across the memory tasks. The results of the present study indicate that persistent memory deficits remain evident in amnestic MCI subgroups using alternate memory tests, suggesting that the concerns regarding potential circularity of logic may be overstated in MCI research.