This chapter focuses on methods for sampling rare populations in health surveys. It begins with a definition of rare populations, considering reasons why it may be necessary to study rare groups, and then gives an overview of some of the issues associated with designing a sample of a rare group. The chapter considers a number of design alternatives, beginning with probability sample methods and concluding with a set of designs that may be characterized as nontraditional or nonprobability, depending upon the implementation. Each of the methods is illustrated with a number of example studies, including a more detailed examination of one or two studies. An overview of traditional, probability methods is also provided. Finally the chapter discusses the methods that are either not probability methods or may be only considered probability samples if certain assumptions can be made.