It has long been known that epilepsy may follow injury to the head. The relationship between the two events hasbeen worked out in some detail for missile wounds of the head by Wagstaffe (1928), Ascroft (1941), and Russell and Whitty (1952), to mention but a few of many such studies, the oppor-tunities for which were provided by the two world wars. The well-defined injury made study precise, and, as the patients became pensioners, follow-up could be effective. The blunt head injuries of civilian life are less easy to study. The numbers involved, the diversity of injury, and the inevitable selection of cases for admission to hospital, pose major problems in obtaining an unselected series in which to study the true incidence of epilepsy and reliable standards to discuss the factors involved. We owe much, however, to the earlier work of Russell (1932), Feinberg (1934), Symonds (1935), Rowbotham (1942), Penfield and Shaver (1945), and Phillips (1954), who have published valuable reviews on this subject. Nevertheless, many of these studies have been on selected cases either due to the population studied, for example Service personnel, or because of the selection involved in admission to a special unit. As Garland (1942) pointed out in a discussion on traumatic epilepsy, the incidence claimed from various sources varied from 20% at its highest down to a figure less than that of epilepsy in the general population. Moreover, there seemed little agreement as to what constituted a mild or severe head injury. Russell (1932) reported an incidence of epilepsy of 3-5% among 200 cases within 18 months ofinjury, Rowbotham (1942) 25% in 450 cases, and Phillips (1954) 6% in 500 cases. In this paper a further attempt has been made to estimate this complication of head injury, based it is believed, on an unselected series of cases. There are several unanswered questions in relation to traumatic epilepsy. What is the significance of epilepsy occurring in the acute