Since the application of self-report measures in criminological research, the correlation between ethnicity/race and crime has become ambiguous and uncertain. To redetermine the presence of this correlation, various researchers have pointed to the need to resolve the methodological dilemma. Some have found common grounds for the various measures of criminality, while other urge that research designs of studies on the criminal justice process be refined to include, for example, longitudinal data covering the various stages of the process, organizational constraints in the criminal justice system, judicial attitudes, and rural-urban differences. Few researchers have recognized, however, the importance of improving the independent variable of ethnicity or race. In most criminological studies, race and not ethnicity has been used. In studies which employ ethnicity instead of race, ethnicity is misleading. Ethnicity is comprised of ethnic identity and ethnic culture retention at the individual level and social organization of the ethnic group at the collective level. The study of the issue of ethnicity and crime thus depends not only on the improvement in methods of ethnicity as an independent variable. A few notes and over 100 references accompany the article. A summary in French is provided.