Obtaining representative samples from populations of emerging adults who do not attend college is challenging for researchers. This article introduces respondent-driven sampling (RDS), a method for obtaining representative samples of hard-to-reach but socially interconnected populations. RDS combines a prescribed method for chain referral with a mathematical model to adjust for potential sampling bias. A case study is presented of the use of RDS to obtain a sample of rural African American emerging adults and to document the feasibility, effectiveness, and efficiency of the method with this population. RDS-related findings indicate that rural African American emerging adults form interconnected social networks that are appropriate for using RDS. RDS procedures are an acceptable and relatively efficient strategy for rural African Americans that is likely to yield more representative samples than traditional snowball sampling and provides data on potential sampling biases for planning multivariate analyses. (Contains 2 tables and 3 figures.)