The challenges involved in the design of ethical, nonexploitative research projects with sex workers or any other marginalized population are significant. First, the size and boundaries of the population are unknown, making it extremely difficult to get a representative sample. Second, because membership in hidden populations often involves stigmatized or illegal behavior, concerns regarding privacy and confidentiality are paramount and difficult to resolve. In addition, they often result in challenges to the validity of the data. Third, in spite of evidence to the contrary, associations between sex work and victimization are still strong, dichotomies remain prevalent, and sex workers are often represented as a homogeneous population. Drawing on three research projects in which the author has been involved - all grounded in a sex-as-work approach - as well as the work of others, this article provides several strategies for overcoming these challenges. Clear guidelines for ethical, nonexploitive methodologies are embedded in the solutions provided.