In the presurgical evaluation of patients with pharmacoresistant epilepsies, the borders between disciplines tend to oscillate. In particular, the necessity to evaluate the functional capacity of a cortical area that has been shown to be epileptogenic and is thus being considered for resection often makes it impossible not to trespass between neurophysiology and neuropsychology. In epilepsy surgery, the principle of “first, do no harm”, often blurs the distinction between “bed and bench”: findings from basic sciences that implicate a certain brain region in specific neurological or neuropsychological functions must be taken into account when an elective resection of a structure is considered – even if resection might relieve the patient from frequent intolerable seizures. That may seem to go without saying and, in fact, several examinations that are used routinely in modern presurgical evaluation are really shortcuts in the path from bench to bed. It may be an excuse that there are still are no generally accepted protocols for examinations such as the intracarotid amobarbital (“Wada”) test or language mapping by electro-stimulation, be the latter intrasurgical or by chronically implanted subdural electrodes. On the other hand, the opposite way “from bed to bench and back” has a famous history in epilepsy surgery: a new area of memory research followed Brenda Milner’s analysis of the famous anterograde amnesia caused by the resection of both hippocampi in Henry Molaison (H.M.) and, eventually, this has led to important clinical applications in both neuropsychology and neurophysiology.