To elucidate possible sources of error in the assessment of drug-resistant epilepsy questionnaires were sent, at least one year after hospital discharge, to 259 patients hospitalized between 1. 1. and 31. 12. 1986 because of epilepsy. Information thus obtained included whether and what seizures had occurred after discharge and what drugs had been taken. Answers were received from 173 patients (67%). They revealed that 18 patients (12 men and 6 women; mean age 29.6 [16-58] years) had, before hospitalization, been wrongly thought to have epilepsy resistant to antiepileptic drugs, significantly more often (P less than 0.05) those with generalized epilepsy (21.2%) than with focal seizures (7.4%) or those with both generalized and focal attacks (9.4%). In five patients increasing the dosage of a drug currently taken stopped the seizures. In 12 patients at least one new drug was introduced, in seven of them as a differentiated choice guided by type of seizure and kind of epilepsy. In 11 of these patients the dosage of other drugs could be reduced or the drug discontinued altogether. In one woman seizures stopped after clonazepam had been discontinued. With two exceptions (mesuximide) success was achieved with just one of five standard antiepileptic drugs or by individually determined combinations of them. These results indicate that even in apparent drug resistance absence from seizures can be achieved in many cases of epilepsy if treatment strictly follows accepted principles.