A review of the data supports the suggestion of Knox (1974) that the sex ratio and incidence of anencephalics vary together within populations, as they are known to do between populations. There seems some doubts, though, whether Knox was correct in hypothesising that the sex ratio of anencephalics varies with their incidence in relation to the dizygotic twinning rate. The pattern of variation seems to suggest that there are two sorts of cause of anencephaly, one of which is environmental and affects predominantly female embryos. The other may be environmental or genetic, and seems to attack the sexes in roughly equal numbers. If this hypothesis were correct, it would carry implications for the testing of specific teratogens.