Purpose: To examine the lateralization utility of preoperative verbal retention in patients with and without bilateral hippocampal atrophy. Methods: The sample consisted of 74 patients with EEG-defined unilateral temporal lobe epilepsy (TLE) who had also undergone volumetric magnetic resonance imaging (MRI). Verbal retention was operationalized by the Logical Memory percentage retention subtest (LM%) of the Wechsler Memory Scale. Patients were divided into groups with (a) bilaterally normal hippocampal volumes, (b) unilateral atrophy, or (c) bilateral atrophy. Two different thresholds (empirically derived vs. normative) were used to lateralize on the basis of LM%. LM% lateralization was then examined by group using X2, sensitivity, positive predictive values, and odds ratios. Analyses were also conducted separately in the subset of patients who were seizure free after surgery. Results: Mean LM% performance was significantly lower in patients with left versus right TLE in the subset with bilateral hippocampal atrophy (p = 0.018) but not in patients with a normal MRI (p = 0.918) or unilateral atrophy (p = 0.087). The odds of a correct lateralization by LM% increased from 1.67 in patients with normal MRI to 36.11 in patients with bilateral hippocampal atrophy. The power of a right and left lateralization prediction by LM% was 100% and 75%, respectively, in patients with bilateral hippocampal atrophy. Similar results were obtained when analysis was restricted to patients who were seizure free after surgery. Conclusions: Preoperative verbal retention as measured by LM% may provide, meaningful lateralization information in patients who are difficult to lateralize via MRI.