SummaryIn an analysis of data on the human sex ratio, comprising some 23,700 births, we have been unable to find any evidence that the determination of sex is not a chance effect. We have tested for variations of the probability of a birth being male between families, and between births; for the existence of a correlation between the sexes of births separated by none, one, and two other births in a family; and whether the continuation of procreation in a family is influenced by the sexes of the children previously born.We compare our result with findings from other samples, and conclude that it would be unwise to draw any general biological inferences.We wish to thank Prof. J. A. Book for permission to use the data. The original tabulation was made on the initiative of the late Prof. G. Dahlberg by Mr E. Lander, to whom we are very much indebted for valuable advice.One of us (A.W.F.E.) is particularly grateful to the Scandinavian Studies Fund of the University of Cambridge for financial help.