This essay assumes that the design and use of surveys is a fundamentally rhetorical act. It provides suggestions for employing and designing health-related surveys intended for research participants who might be characterized as inhabiting one or more precarious positionalities. We use “precarious positionality” to signal when research participants self-identify as one or more of the following: a racial and/or linguistic minority, economically disadvantaged, disabled, former or current drug user, undocumented, un(der)educated, oppressed, sexualized, disenfranchised, criminalized, and/or colonized. Drawing on the research team’s experiences with piloting what we hope will eventually become a nationwide survey, the essay describes how to avoid several survey-design pitfalls; it also makes recommendations for how to improve survey-based health research that enrolls participants who inhabit one or more precarious positionalities. Our recommendations attend to rhetorical complexities related to survey ethics, inclusion criteria, privacy, stigmatized and misleading language, variations in discursive repertoires, accessibility, and liability.