Live human births are usually more than half male, in spite of excess losses of males throughout fetal development. These observations together demand an excess of males near the beginning of pregnancy greater than that seen at birth. Reductions of the usual excess of males among human live births have widely been considered to represent consequences of untoward circumstances surrounding conception. Repeated competent research efforts have found no evidence for any bias in gametogenesis or fertilization in favour of Y-bearing sperm. Male embryogenesis is faster and more efficient, leaving females in excess among failures before the fetal period. Sex differences in speed and efficiency of embryogenesis, dependent for example on epigenetic differences such as genomic imprinting, produce an excess of males at the transition from embryogenesis to clinical pregnancy, that will survive the male excess of losses throughout the fetal period, to yield an excess of males among live births. Changes in, or mediated by, the epigenetic environment of embryogenesis provide the most plausible prospects for causes of changes in secondary sex ratio.