The Civil Rights Act prohibits not only discriminatory treatment but also discriminatory effect, i.e., the unintentional exclusion of a disproportionately large number of members of a minority group unless the exclusion is justified as being job related. Since minorities suffer a disproportionately large number of arrests and convictions, employers' use of these records is often subject to the job-relatedness standard. However, available survey data suggest that employers (especially in large companies) collect and utilize criminal record information in ways prohibited by Title VII. The disparity between legal provisions and employers' practice is partly due to court disagreement on how disparate impact should be measured. The article argues that plaintiff-job applicants should be deemed to have demonstrated that a criminal record selection has an adverse impact if they can show that the criterion would negatively affect a significantly greater percentage of minorities than nonminorities among the regional populace. White employees can also assert disparate treatment if they can show that in a previous or simultaneous incident a minority applicant or employee had received preferential treatment. Though courts sometimes admit even arrest records as screening devices in case of job-related arrest, arrest records prove in no way that the individual was actually involved in the crime and therefore, should not be a factor in employment decisions. The article recommends that the Equal Employment Opportunity Commission lead the effort in clarifying legal standards and producing guidelines on the employment-related use of criminal records. Case notes are cited. (Author abstract modified)