Clinical observation of performance on the Logical Memory (LM) and Visual Reproduction (VR) subtests from the WMS-III has revealed some variability in retention rates across stories and figures. This paper examined the degree to which this variability occurs in lateralized temporal lobe epilepsy (TLE) in comparison to a matched group from the WMS-III standardization sample, and explored whether analysis of qualitative aspects of LM and VR performance yield additional lateralizing information in TLE. Analysis of LM and VR scaled scores revealed differences between the TLE groups for LM, but not VR scores. All subjects benefited from repetition of LM Story B, with greater improvement in story retention in the Left versus Right TLE group. Variability in VR recall across figures was seen in all groups, with a bimodal distribution of retention rates for each figure and a sizable percentage of each group completely forgetting two or more figures. These results suggest that more careful analysis of individual LM story performance may be useful in some patients with TLE, whereas variability in VR retention across figures is common and should not be over interpreted.