It is clear that alcohol use, though pervasive among offender groups does not cause crime. Therefore, attempts must be made to identify how alcohol interacts with other moderating factors that result in criminal activity. Relevant research in this area is reviewed, with a special emphasis on methods of data analysis, choice of subject population, use of self-report data, interaction of type of crime and alcohol consumption patterns, victim intoxication, multiple substance abuse, and demographic factors. More sophisticated methodologies and more precise definitions of both type of crime and type of alcohol abuse appear warranted. Examples of innovative approaches to identifying and studying nonincarcerated criminal populations are suggested as ways to further our understanding of the alcohol-crime link. In addition, the use of mechanical/chemical tests for determining blood-alcohol levels are advocated as a way of increasing the validity of self-report data. 47 references. (Publisher abstract)