Four surveys were used in the analysis: (1) nationally representative survey of runaway and homeless youth residing in federally and nonfederally funded youth shelters; (2) multicity survey of street youth; (3) nationally representative household survey of youth with and without recent runaway and homeless experiences; and (4) nationally representative household survey of youth whose previous runaway and homeless experiences were unknown. Comparable measures of alcohol, tobacco, and drug use were included in all four surveys, although not all measures were available from each survey. Results showed runaway and homeless youth used substances more than youth living in households. Differences in prevalence rates were especially pronounced for illicit drug use. Findings also demonstrated the risk of substance abuse for runaway and homeless youth varied, depending on current living circumstances. Street youth had markedly higher rates of substance use than shelter youth and those living at home who had run away or been homeless in the previous year. Substantially more street than shelter youth used intravenous drugs, heroin, methamphetamines, and crack cocaine, drugs associated with high risk and high dependence. Street youth were also involved in more serious drug use, as indexed by stage of drug involvement and the number of substances used. Youth in households with recent runaway experiences had higher prevalence rates of alcohol use than shelter youth, although prevalence rates of illicit drug use were similar. Limitations of the study findings are noted, and suggestions for improving drug intervention and treatment services for runaway and homeless youth are offered. 27 references, 3 tables, and 1 figure