This report describes efforts to develop sampling methods to be used in national or regional studies of migrant and seasonal farm workers (MSFWs). Several facets of the MSFWs' lifestyle create sampling difficulties. One is mobility. Although the dynamic nature of MSFWs' movement is partly understood, it is sufficiently unpredictable to create problems, such as "multiplicity"--multiple opportunities for the same individual to be selected. Other problematic aspects relate to identifying all housing locations of migrants in a given area, the transitional periods between migrants' jobs, variations in the timing of growing seasons around the country, and the elusiveness of undocumented aliens. The proposed design involves a stratified four-stage cluster sample. The first stage uses area sampling methods for small groups of neighboring counties, while other stages use methods of list sampling for enumerating districts or block groups, migrant camps, and households. Sampling units, stratification, selection, and methodological issues are discussed for each stage. A large part of this document consists of Rebecca Robin Parker's 1989 master's thesis on this topic, which provides details on: (1) important migrant worker datasets compiled by the Migrant Student Record Transfer System, Douglas Massey, and Richard Jones; (2) definitions and characteristics of migrant farmworkers population; (3) multiplicity and how to handle it; (4) migration patterns; and (5) a description of a proposed 4-stage sampling design. An appendix contains a related paper by William Kalsbeek that discusses conceptual framework and design parameters and provides statistical formulas. (SV)