Statistics from around the world have indicated that about half of all sexual offenses are committed by a stranger to the victim. Investigating stranger sexual offenses is particularly difficult; researchers have devised classification systems that attempt to differentiate between types of sexual offenders in order to inform offender treatment, risk of recidivism, and criminal justice decisionmaking. These classification systems can also be useful to police officers seeking to prioritize the field of suspects in stranger sexual offenses. The authors review and critique early clinical classification systems that displayed a lack of empiricism before turning their attention to Massachusetts Treatment Center (MTC) classifications, which they contend are empirical but are inappropriate for criminal investigative use. Criminal profiling within the Federal Bureau of Investigation (FBI) is similarly reviewed as showing progress but lacking in scientific rigor. An outcome of the criticisms leveled against FBI criminal profiling has been the development of two empirical techniques for criminal classifications: item-to-item correlation analysis and thematic behavioral analysis. Each of these is described and critiqued before the authors move on to a discussion of new approaches to suspect prioritization that are informed by the complementary frameworks of pragmatism and research in naturalistic decisionmaking. The difficulty of constructing offender classification systems that are useful to police investigations underscores the need for academics and practitioners to collaborate on the construction of offender classification systems. References