Old age is recognized to be the commonest time in life to develop epilepsy. There is a perception that older patients are more sensitive to the deleterious cognitive effects of antiepileptic drugs (AEDs). Elderly patients (median age 70 years, range 60-88 years) taking anticonvulsant monotherapy (10 carbamazepine [CBZ], 8 sodium valproate [VPA], 5 phenytoin [PHT]) took an extra dose of their usual medication (200mg CBZ, 500mg VPA, 100mg PHT) and matched placebo each for a month in random order. The concentrations of AEDs were higher after 7 and 28 days of active treatment compared with placebo (7 days: CBZ 9.5 vs. 7.8 mg L(-1), p < 0.05; VPA 97 vs. 64 mg L(-1), p < 0.05; PHT 13 vs. 11 mg L(-1), p < 0.05; 28 days: CBZ 9.4 vs. 7.7 mg L(-1); p < 0.01, VPA 85 vs. 60 mg L(-1), p < 0.05; PHT 16 vs. 13 mg L(-1), p < 0.05). Despite these increases in concentration, there were no significant changes in attention, reaction time, finger tapping, memory, side-effect scale or sedation scoring during the active phases compared with placebo phases for the three drugs analysed together and separately. Elderly patients taking standard AEDs as monotherapy did not develop cognitive impairment when the dose was modestly increased within the target range for each drug.