Since the 1950’s, the Philippines, particularly the National Statistics Office (NSO) have employed the concept of a master sample (MS) in the design and conduct of its household surveys. A master sample as used in this context is defined as a sample from which subsamples can be selected to serve the needs of more than one survey or survey rounds (UN-NHSCP). The master sample consists of a randomly selected set of geographical areas, termed primary sampling units (PSUs), which are taken to be barangays or combinations of barangays. The samples of households and persons for all of the surveys are then selected within that set of PSUs. The use of a master sample of PSUs has the benefits of using the same set of interviewers for each of the surveys and of sharing the costs of updating listings of housing units within
the PSUs across several surveys. This paper documents all the important processes that were followed in designing the 2003 Master Sample (MS). In particular, this paper provides a description of the choice of the primary sampling units, determination of sample size in terms of the number of housing units/households, the choice of domain, selection procedures adopted, weighting and estimation procedures, and rotation of samples.