Attempts to evaluate the deterrent effects of criminal sanctions are frustrated by four main obstacles: scarcity of evidence, interpretive problems, methodological defects, and theoretical inadequacies. For over a decade sociologists and criminologists deliberately discounted the value of force and coercion as crime deterrents, concentrating, instead, on socialization, structural integration, or value consensus. Hence there is a lack of evidence in favor of sanctions. Additionally, interpretive problems have precluded the proper use of even the few data that have been gathered. Criminologists disagree on the meaning of the term deterrence, for example. Studies on general deterrence reveal the difficulties inherent in trying to assess this variable, even by means of self-report surveys. Specific deterrence (i.e., the effect of sanctions on the behavior of the offenders who have been punished) is also hard to evaluate since rearrest and recidivism statistics are not systematically compiled. Theoretical inadequacies also hinder useful evaluation. In spite of these obstacles, evidence sufficient to support the deterrence hypothesis does exist. Researchers should focus on the conditions under which sanctions have various effects; moreover, evaluators should broaden the focus of their inquiry to encompass other plausible outcomes of criminal sanctions besides deterrence, such as brutalization and prisonization, which may increase crime instead of reducing it. Researchers and evaluators must also avoid the mistakes that have so far plagued inquiry, especially ideological bias in research. A list of 77 references is appended.