The article opens with a brief review of the development of modern public-opinion surveying on crime and punishment, particularly research on crime seriousness and judgments of appropriate punishment. Some of the findings of this research are highlighted. This is followed by a consideration of the various reasons why suspicion can reasonably be cast on the findings of this research and a discussion of some possible explanations for the findings in the context of these suspicions. The article concludes that underlying the research on public sentiment is the assumption that the public has a discrete and discernible body of sentiment that can be accessed through the survey techniques currently available. An analysis of survey methodology that has identified public sentiment regarding judgments of punishment for crime indicates that faith in the existence of public opinion regarding policy-related issues in crime and justice may be premature. There are a number of problems that call into question the validity of the survey data. Respondents are sometimes asked to make technical judgments about complicated crime and offender situations that involve numerous variables. The complexity of these situations, in conjunction with the time and demand constraints of the telephone interviews used in many surveys, raises serious questions about what the responses represent. 35 references