Congenital malformations of the eye can cause blindness in children. They occur throughout the world and in most cases the aetiology is unknown. Linkage studies have largely been unsuccessful and the risk to siblings is generally low. Epidemiological and laboratory evidence support a hypothesis that there may be genetic (recessive) predisposition to the teratogenetic effects of mild to moderate maternal vitamin A deficiency (VAD) during pregnancy. This may explain the higher prevalence of congenital eye anomalies in a part of Asian countries, where maternal VAD is common and consanguineous marriages are popular. Other congenital malformations commonly found in association with ocular coloboma (e.g. oesophageal fistulae and heart defects in CHARGE association) may also be VAD related. Mutations in a gene involved in the cellular access to vitamin A that normally protects the tissue or embryo from natural variation in dietary vitamin A intake, could render that individual intolerant of conditions of VAD. An interaction of this kind could also explain a proportion of "sporadic" cases in locations where VAD is uncommon. If this interaction is shown to be true, there are public health implications for the prevention of blindness due to congenital eye malformations. The hypotheses proposed above are reminiscent of the research leading to the discovery that folic acid supplementation could prevent neural tube defects. However, this form of intervention would be much more difficult with vitamin A, which is itself a powerful teratogen if present in excess.