1. Data on more than 2 1/2 million twin births suggest that the regression of sex ratio in twins on maternal age does not decline monotonically like that of singletons, but, like the incidence of dizygotic twinning, seems to rise and then fall with maternal age. 2. Accordingly it is hypothesized that the sex ratio in monozygotic twins is lower than that in dizygotic twins or that in singletons. This would account also for the low overall sex ratio in twins. 3. The data are consistent with the hypotheses that the monozygotic twin sex ratio is constant for all maternal ages at a value of about 0.496, and that the dizygotic twin maternal age-specific sex ratios are the same as the singleton sex ratios for the same maternal ages. 4. The hypothesized low sex ratio in monozygotic twins is reminiscent of that in some congenital malformations: possibly some aetiological factor is common to monzygotic twins and such congenital malformations. 5. It is suggested that the total-birth sex ratio of triplets is underestimated in national vital statistics records.