PURPOSE: To compare hippocampal and neocortical metabolite ratios using single-voxel proton MR spectroscopy with different echo times in patients with complex partial seizure. MATERIALS AND METHODS: Using a GE Signa 1.5 T scanner with STEAM and PRESS sequences, automated single voxel proton MRS was used to determine metabolite ratio differences in the hippocampus and neocortex of nine complex partial seizure patients [mesial temporal sclerosis (n=5), status epilepticus (n=1), tumor (n=1), cortical dysplasia (n=1), occipital lobe epilepsy (n=1)]. A total of 20 examinations were performed in the region of the hippocampus (n=17), temporal neocortex (n=1), and parieto-occipital gray matter (n=1). Voxel size range was 5.2 -17.4 cm 3. The calculated creatine (Cr) peak was employed as an internal reference and the relative ratio of N-acetylaspartate (NAA) and choline (Cho) was calculated for both short and long echo times using an automated PROBE/SV (GE Medical Systems) package. Each NAA/Cho ratio obtained using both PRESS and STEAM techniques was compared by means of statistical analysis (paired Student t-test). RESULTS: Using PRESS (long TE, 272 ms), NAA/Cho ratios were successfully calculated in 16 of 20 examinations; in four this was not possible due to noise levels of the Cr and Cho peaks. Using STEAM (short TE, 30 ms) NAA/Cho ratios were successfully calculated in 19 of 20 examinations; in one, the Cho peak could not be measured. Using PRESS and STEAM, mean and standard deviations for the NAA/Cho ratio were 1.22 +/-0.50 and 1.16 +/-0.36, respectively. There were no statistically significant differences in this ratio between the short and long TE method (p <0.01). CONCLUSION: In complex partial seizure patients, no significant metabolite differences were found between short and long echo times of single voxel proton MR spectroscopy. The metabolite ratio at different echo times can be reliably obtained using this simplified and automated PROBE/SV quantitation method.