Fifty-five patients with intractable partial seizures whose on-medication EEGs demonstrated either predominantly focal epileptiform lesions or absence of paroxysmal activity, were studied, and the effect of withdrawing all anticonvulsive drugs on their EEGs was observed. Four types of response were encountered: (1) no effect (20%); (2) specific (focal) activation (25%); (3) complex activation (29%) with widespread of the initial on-medication focus or appearance of the additional independent epileptogenic foci; and (4) "non-specific" activation (63%), consisting of bursts of either bilaterally synchronous and frontally dominant spike and waves, triphasic waves, or sharp slow complexes, or smaller amplitude rapid and diffuse spike-and-wave complexes. This latter effect is thought to be secondary to metabolic derangements resulting from the withdrawal of neurotropic agents and not directly related to the specific epileptogenic process. No association was found between type of effect and any of the following parameters: topography of on-medication focus, duration of therapy, type of anticonvulsant used, suspected underlying etiopathology, or median age when medication was withdrawnn. Furthermore no evidence could be found that the development of a "complex" or "non-specific" EEG effect carried with it a bad prognosis for surgical cure following focal cortical excision. Performing off-medication tracings seems to be of greatest value in patients with partial seizures and EEGs revealing either a relative paucity of definite absence of epileptiform discharges. The occurrence of a "non-specific" response in a questionable epileptic during the off-medication period, on the other hand, should be interpreted with caution.