Discusses investigations of allegations of child mistreatment, especially cases that are screened out. Fewer child abuse or neglect reports may not indicate an absolute decrease in child mistreatment, but rather mask deficiencies in screening practices or result as an unintended consequence of welfare reform (e.g., parents leaving children inadequately supervised in order to meet work requirements). Few states have explicit guidelines describing the types of reports that should be screened out, levels of review differ across states, and training requirements also vary. Quantification of child mistreatment allegations revealed that New Jersey screened out the least (5 percent of cases) and Vermont screened out the most (78 percent). Moreover, it is possible that states or jurisdictions with heavy case loads will screen out a large portion of reports and also fail to substantiate mistreatment in those cases that are investigated, in an effort to offset overwhelming demands. Effective screening could(1) reduce caseloads, thereby making the best use of limited resources and (2) reduce unnecessary intrusion into families’ lives by weeding out inappropriate reports. Further research is needed to address concerns about missing true cases of mistreatment and to assess the frequency of re-reporting and substantiation of cases initially screened out. (22 references)