The authors examined clinically and electrographically 100 patients with severe stenosis of the internal carotid artery. The examination was made before and after carotid endarterectomy (CEA). In the scalp SEP response of the median nerve the authors evaluated the modulation of the structure of the response at the period of the 15 th – 100 th (200 th ) ms and latency and amplitude of the component of the thalamocortical entry on the CEA side and on the other side. The responses were synchronously recorded from the central C3/C4 area with reference to Fz and A1/2 and from the frontal area (approximately F3/F4) where scalp projection of the prerolandic somatosensory generator is assumed. After surgery the architecture of the response and latency did not change significantly. On evaluation of the whole group no significant differences were recorded in amplitudes but in 12% of the patients after operation the amplitude of N20/P25 increased by more than 100%. This increase was (with a single exception) not associated with improvement of the clinical condition – the authors evaluate it as a manifestation of a relevantly increased activation of the cortical somatosensory generator with subclinical impact. SEPs P22/N30 from a prerolandic lead in 8% were not differentiated at all, and in 34% vaguely. Their absence (imperfect differentiation) was not associated with hemiparesis nor with CT/MRI abnormality in the frontal lobe. In 58% the SEPs P22/N30 were differentiated, in relation to N20/P25 they were architectonically equivalent and hastened by 0.1–2.0 ms. The spatial time relation- ship between N20/P25 and P22/N30 provides evidence that P22/N30 develops rather by a mechanism of a volume conduction, spreading from the central region, than by a discharge in a special frontal generator. In the diagnosis of cerebral ischaemic disorders the examination by scalp SEPs P22/N30 was no asset.

        Key words: carotid endarterectomy (CEA), SEPs of median nerve, P22/N30