We have compared frequency and severity of epileptic seizures during pregnancy with the 9 months prior to pregnancy in 66 patients with a total of 78 pregnancies. Data on total number of seizures and major seizures are analysed separately. No statistically significant differences between frequency before and during pregnancy were found. Cases with seizures before pregnancy tended to have seizures during pregnancy, but there was no association between occurrence of seizures prior to pregnancy and increased frequency during pregnancy. There was no evidence that seizures became more severe during pregnancy. No relationship was found between type of epilepsy and change in seizure frequency during pregnancy. The data suggest that our therapeutic interventions may have influenced seizure frequency in at most a minority of cases. Serum concentrations of the anti-epileptic drugs were monitored regularly during pregnancy. Only data on patients on constant drug dosages and, therefore, presumedly with the mildest seizure disorders were analysed. In the majority of cases there was a decrease in drug levels during pregnancy. Most cases tolerated this decrease without an increase in seizure frequency. On the other hand, most cases with increased frequency also had decreased drug levels.