OBJECTIVE: To test pathophysiologic hypotheses regarding the occurrence of a splenial lesion in patients with epilepsy. METHODS: The authors studied 16 patients with a splenial lesion and 32 control patients, all of whom had MRI examination immediately after presurgical EEG long-term monitoring (LTM). The authors compared the number of generalized tonic-clonic and partial seizures during LTM, antiepileptic drug (AED) withdrawal, and laboratory results. RESULTS: All of the patients with a splenial lesion had their AEDs stopped completely, vs 47% of the controls (p = 0.001). Patients with SCC lesion had a longer duration of complete withdrawal (median 3.5 vs 2 days, p = 0.03). There was no correlation with seizure frequency or the introduction of new AEDs. CONCLUSION: A lesion of the splenium of the corpus callosum in patients with epilepsy is not associated with toxic drug effects or high seizure frequency, but might be induced by a rapid and relatively long-lasting reduction of antiepileptic drugs. Its frequency might be underestimated as MRI after long-term monitoring is rarely done.