This paper outlines the processes that were undertaken in the development of the master sample, namely: the construction of the sampling frame, the determination of sample size and sample allocation across domains; the stratification of the primary sampling units (PSUs) and the allocation of the sample across PSUs and the selection and implementation procedures of PSUs and households for Bhutan, one of the most isolated but the happiest least developed country in the world that is sandwiched between the People's Republic of China and India. The resulting master sample provides a more efficient set of PSUs with survey weights that do not vary very much within domains. The master sample is also expected to provide equally efficient estimates for the two classes of estimates discussed above – for subgroups that cuts across domains and for estimates at the domain level.