In the first phase of the project, a network sampling (NS) procedure was evaluated in which respondents were asked to report about their own household, neighbors, and coworkers. Respondents included 600 random-digit-dialing households and 300 households selected from police missing child reports for the Chicago Metropolitan Area. NS uncovered substantially greater number of missing children than found by direct screening. Relatives were the major source of additional information, followed by next-door neighbors. Almost all reports involved runaways. Agreement between NS and household reports was very low, but most of the discrepancy was for relatively less serious, short-term incidents. Results suggest that while NS should not be used to locate or estimate the number of runaways, it should be considered for estimating or locating child abductions. In the second phase of the project, the feasibility of using the capture/recapture (CR) technique for estimating the number of runaways/throwaways was studied in Illinois and District of Columbia samples. Results indicate that CR was not very effective for runaways and throwaways staying with relatives or friends, but may be effective in determining thoses who are out on the street with no fixed place to stay. Appendixes provide supplemental research information and documents. 21 tables, 1 figure, and 24 references.