CT, MRI, 18/FDG-PET and Depth. EEG, performed with subdural and depth electrodes were part of the presurgical evaluation in 22 patients. Statistical analysis of 18/FDG-PET was performed to compare cerebral utilization of glucose to that of normal age matched controls. The findings of CT, MRI, and quantitative analysis of PET are compared with those of ictal Depth. EEG. A positive correlation between CT and Depth. EEG was obtained in 23% of the patients and between MRI and Depth. EEG. in 50%. For both imaging techniques a negative correlation was found in 5%. Regional abnormalities were found with quantified PET in 95% of the patients and were concordant with Depth. EEG. for side of onset in 77% of the patients and for lobe of onset in 59%. A possibly false localising PET result for lobe of onset was obtained in 8 patients (36%). Limitations of PET were most apparent in patients with regional mesiolimbic or bilateral seizure onset. A favourable outcome of surgery was associated usually with positive convergence of both methods. PET may be a valuable contribution to the research and management of partial complex epilepsy, but at present cannot be considered a reliable alternative to invasive EEG methods in patients without clear unilateral focus localization on surface EEG.