Editor—T Eke and colleagues report three cases of severe, symptomatic constriction of the visual fields associated with vigabatrin treatment.1 We have had experience of a similar case. A 34 year old man who had had refractory partial onset seizures since the age of 8 was taking vigabatrin 3000 mg daily (since 1989), carbamazepine 1200 mg daily, and sodium valproate 5000 mg daily (both since 1982) when he suddenly developed visual deterioration with blurring and loss of peripheral vision in July 1995. Surface electroencephalography suggested a left hemispheric focus, but magnetic resonance imaging of the brain showed no abnormality. On examination he had impaired visual acuity, with corrective lenses to 6/9 in both eyes. Pupillary responses were normal and confrontation visual fields full. There was pronounced bilateral optic atrophy and a maculopathy, both more evident in the right eye. His electro-oculogram was flat, and he had subnormal cone and rod electroretinograms. Visual evoked responses were normal. Vigabatrin was stopped, but his damaged retinal pigment epithelium and photoreceptors did not improve. Vigabatrin was licensed in Britain and the Republic of Ireland in 1989; these were its first markets.2 The mechanism by which the drug might produce retinal damage is unknown, and the symptoms did not improve in this patient when the drug was withdrawn. In addition, all of the patients reported on were taking other antiepileptic drugs at presentation. Our own patient had been treated previously with phenytoin, primidone, and lamotrigine. Chronic refractory epilepsy may also have a role in causing these unusual lesions. For example, disseminated intravascular coagulation associated with prolonged seizure activity may have been responsible,3 rather than a toxic effect of lamotrigine. Although it is reasonable to recommend ophthalmological review in patients with visual symptoms taking vigabatrin, we are still a long way from establishing a …