Safety education directed at parents of pre-school children may be an important tool in reducing injuries of pre-school children at home. However, many studies on educational programmes directed at parents have shown little evidence of success in terms of changes in parental safety behaviour and childhood injuries. These programmes may have been affected, among other factors, by a typical kind of 'programme failure' due to a lack of empirical information on at-risk behaviour and the deterinants of that behaviour prior to the educational development of programmes. We tried to get an indication of the extent to which previous safety educational programmes had grounded their activities on an empirical basis. Therefore, a review of literature was carried out on empirical studies on (i) parental safety measures-the form of parental safety behaviour that is supposed to be very important in preventing injuries to pre-school children-and (ii) the behavioural determinants of those safety measures. It appeared that information on these topics is scarce and this suggests that previous programmes may indeed have suffered from such programme failure. Recommendations are made for enlarging the body of knowledge essential for the development of effective safety education directed at parents of pre-school children. Finally, some recommendations are presented for safety practitioners.