Editor,—While Hoyt and Fredrick rightly state the heterogeneous aetiology of cortical visual impairment (CVI) in children,1 we were disappointed in their bland and unjustified dismissal of the value of electrophysiological studies in these cases. They fail to quote any electrophysiological studies more recent than 1979 in their review, of which there have been a not inconsiderable number. As for our own work we have found that a normal flash VEP indicates statistically and clinically a better prognosis in blind babies with non-ocular visual impairment even though we would concede that in some cases flash VEPs can be normal in CVI.2 It is therefore clinically useful …