Poverty is a significant problem in rural America. Gaining access to economically marginalized rural populations in order to recruit individuals to participate in a research study, however, is often a challenge. This article compares three different nonprobability sampling techniques that have been used to recruit rural, low-income mothers--purposive sampling, respondent-driven sampling, and mixed purposive sampling. We review the relative advantages and drawbacks of the three methods in terms of access to the targeted population, methods of recruitment, size of the sample pool, randomness of the sample, generalizability of results, and researchers' control over the process. (Contains 1 figure, 3 tables and 8 footnotes.)