The Millennium Development Goals (MDGs) triggered increased demand for data on child and maternal mortality for monitoring progress. With the advent of the Sustainable Development Goals (SDGs) and growing evidence of an epidemiological transition towards non-communicable diseases, policy makers need data on mortality and disease trends and distribution to inform effective policies and support monitoring progress. Where there are limited capacities to produce national health estimates (NHEs), global health estimates (GHEs) can fill gaps for global monitoring and comparisons. This paper draws lessons learned from Thailand's burden of disease study (BOD) on capacity development for NHEs, and discusses the contributions and limitation of GHEs in informing policies at country level. Through training and technical support by external partners, capacities are gradually strengthened and institutionalized to enable regular updates of BOD at national and sub-national levels. Initially, the quality of cause of death reporting in the death certificates was inadequate, especially for deaths occurring in the community. Verbal autopsies were conducted, using domestic resources, to determine probable causes of deaths occurring in the community. This helped improve the estimation of years of life lost. Since the achievement of universal health coverage in 2002, the quality of clinical data on morbidities has also considerably improved. There are significant discrepancies between the 2010 Global Burden of Diseases (GBD) estimates for Thailand and the 1999 nationally generated BOD, especially for years of life lost due to HIV/AIDS, and the ranking of priority diseases. National ownership of NHEs and effective interfaces between researchers and decision makers contribute to enhanced country policy responses, while sub-national data are intended to be used by various sub-national-level partners. Though GHEs contribute to benchmarking country achievement compared with global health commitments, they may hamper development of NHE capacities. GHEs should encourage and support countries to improve their data systems and develop a data infrastructure that supports the production of empirical data needed to underpin estimation efforts.