Consistency was assessed by comparing the data on a sample of incidents that were likely to involve a non-family abduction of a child. Findings show that the completeness of the Federal data was reflected by the proportion of local incidents available for analysis at the Federal level. Overall, 88 percent of all incidents reported to local law enforcement agencies were available for analysis at the Federal level. In four of the five agencies studied, analysis showed that to a high degree the incident characteristics found in the local IBR (incident-based reporting) systems were also found in the Federal data. The independent IBR coding by local and State personnel of the paper files from one local agency provided an opportunity to assess the impact of inter-coder reliability on the quality of the NIBRS data. The low consistency rate of the data from this one agency at the local and State level raises serious concerns about the effect of inter-coder reliability on the Federal NIBRS data. Although there is evidence from a study of sample attrition that about 12 percent of incidents never find their way into the FBI data file, updated incident information developed after the initial transfer of data files appears to have consistency rates that are relatively high. Incompatibly designed system coding structures at the local or State levels may result in inconsistent reporting of information. Data transformation procedures at the local or State level may also affect the reliability of data at the Federal level. Four recommendations are offered to increase the ability of the Federal NIBRS data to support research on the non-family abductions of children. 3 tables