The long-term high sex ratio at birth (SRB) is a serious issue in China. In this study, changes in SRB were decomposed into variations in SRB by birth order and compositional changes in female births by birth order. With SRB data from China’s surveys and censuses, and SRB data from South Korea’s vital registration and censuses from 1980–2015, the trend and decomposition results in SRB were compared between China and South Korea, and the decomposition results for urban and rural SRBs, and for provinces, are presented. In both China and South Korea the rise in the SRB was driven by a rise in the SRB at all birth orders, which was only partly counteracted by the change in the distribution of births by order. The overall rise in the SRB ended when there was a decline in the SRB at second birth or above in South Korea. In China the total effect of variations in SRB of all birth orders increased more for the rural population than for the urban population before 2000, resulting in a higher total SRB for rural than urban population. After 2000, the total effect of variations in SRB of all birth orders lowered the total SRB for the rural population, whereas the effect of compositional change increased the total SRB, leading to a very slight rise in the total SRB for the rural population. At the province level, there was no spatial autocorrelation for the changes in total SRB by province, the total effect of variations in SRB of all birth orders or the effect of compositional change. The effect of variations in SRB by birth order accounted for the majority of changes in total SRB in most provinces.