Abstract
This paper describes the feasibility, validity, and cost‐effectiveness of supplementing surveys of shelter clients with surveys of people from nondwelling places (e.g., public places such as bus stations, parked cars and trucks, and abandoned buildings), nonresidential service locations (e.g., soup kitchens, day programs, health clinics), and other residential settings (e.g., jails, hospitals, group homes). The purpose of these supplements is to develop a more accurate estimate of how many homeless people there are, what their needs are, and the extent to which their needs are being met. The essential feature of these attempts involves systematic searches on a sample basis for homeless persons in these places. Reviewing ten studies that have attempted such surveys in the past decade and four that are now attempting to do so, this paper examines the comparative advantages and disadvantages of each approach and when each would be most appropriate. Particular attention is paid to the feasibility and desirability of surveying nondwelling or street locations.

This publication has 3 references indexed in Scilit: