Given a sample--that is, a specified selection of individuals from whom specified information is desired--two questions immediately arise: How much effort should be spent toward getting complete coverage of all individuals in the sample? What methods can be used to adjust for non response, and how successful are such methods? Although only partial answers to the nonresponse problem can be given, this review of techniques and the presentation of a few concepts and results relating to the nonresponse problem should be helpful, particularly to persons who are planning surveys.