Quantitative evidence of hippocampal atrophy has been correlated with site of seizure onset, hippocampal neuronal loss, and seizure relief after resection. Most studies have quantified hippocampal atrophy using ratios or differences between right and left hippocampal values. However, bilateral hippocampal atrophy may remain undetected by these techniques. To assess the frequency and implications of bilateral hippocampal atrophy, we studied absolute hippocampal volumes in 53 temporal lobectomy patients who had undergone intracranial electroencephalogram recordings preoperatively. Coronal images were constructed perpendicular to the longitudinal axis of the hippocampus. Atrophy was defined as > 2 SD below control values in the volume of the posterior 1.5 cm of the hippocampus. Five of 53 patients (9%) had bilateral hippocampal atrophy; four of these cases were undetected by ratios. Surgery was performed on the side of ictal onset in all five patients; four have been seizure-free for > 2 years. These results suggest that (a) mesial temporal sclerosis can be present bilaterally and may go undetected by hippocampal ratio or difference measures; (b) absolute hippocampal volume values as well as ratios are needed to detect all patients with bilateral hippocampal atrophy; and (c) temporal lobectomy is not contraindicated in patients with bilateral hippocampal atrophy, but success depends on electroencephalographic documentation of the side of predominant ictal onset.