It is crucial to monitor global health indicators accurately in order to optimize resource allocation. This is especially so in developing countries where the improvement of these health indicators are most needed. However, estimating and validating these indicators are fraught with challenges, one of which being the paucity of accurate data. The Bayesian modeling approach implemented in this thesis provides more objective, data-driven insights into estimation of global health indicators. We provide a set of important analyses and fill the previous research void on selected indicators. We take account of the data quality that varies across different sources as well as infer the levels and trends of indicators in countries and periods with limited data by data-rich country-years. The resulting estimates provide new insights into child and maternal mortality and the sex ratio at birth globally. The methods and results can be used by international agencies for policy making.