An introductory review is given of some general aspects of the genetic basis for the determination of sex, with particular reference to vertebrates. The Weismann-Fisher view is that sex arose because it was of selective advantage. The bisexuality of vertebrates is not necessarily ‘best’ for the group in terms of selective advantage; however, the very simplicity of the bisexual system may have been advantageous. The common near-equality in the numbers of the sexes is explained by the Darwin-Fisher ‘ automatic 1:1 tendency’, and Fisher’s concept of parental expenditure can explain deviations from equality and also changes in the ratio during embryogenesis. A direct selective advantage can be argued for the 1:1 ratio—or, more generally, for that ratio which maximizes the chance of a successful encounter of the sexes. Possible modes of gene action affecting sex ratio are discussed. A new assessment is made of the identity of the heterogametic and homogametic sexes in vertebrates, the conclusions for Amphibia deviating from those in some of the literature. It is maintained for vertebrates in general that the ‘ odd ’ sex chromosomes (Y or W) of a species determine strongly in the direction of one gonadal sex. Responsibility for determining the opposite sex must lie between the autosomes and the ‘ even ’ chromosomes (X or Z), but there is little evidence that the ‘even’ chromosomes are specifically implicated. Sex differentiation may not be controlled by a limited number of sex-determining genes, and the ultimate units may be aggregates of the genetic material larger than genes. The possibility of controlling sex ratio at conception by altering the proportion in which X- and Y-bearing spermatozoa fertilize eggs is discussed in relation to the more general question of whether the phenotype of a developing or mature spermatozoon can be affected by its own haploid genetic content.