The public sectors of care included alcohol and drug, juvenile justice, mental health, public school services for youths with serious emotional disturbance, and child welfare. Study participants were 936 adolescents 13-18 years old who were randomly sampled from all youths who were active in at least one of these public sectors of care in San Diego County, CA. SUD's were assessed through structured diagnostic interviews conducted between October 1997 and February 1999. The Personal Experience Inventory (PEI) was used as a screening tool to identify adolescents who had used any substance. Youth found to have used a substance were administered the Composite International Diagnostic Interview-Substance Abuse Module (CIDI-SAM) to diagnose tobacco, alcohol, and marijuana use disorders and individual diagnoses for other illicit substances. Logistic regressions were used to examine relative prevalence rates by sector. For lifetime diagnoses, 39.5 percent of the sample met criteria for at least one SUD. Asian/Pacific Islander and Latino youths did not differ significantly from Caucasians in prevalence rates for any lifetime SUD or for alcohol, cannabis, stimulant, or opiate use disorders. African-American youths, compared to Caucasians, reported significantly lower rates of any SUD, alcohol, cannabis, hallucinogen, and stimulant use disorders. Latino youths were significantly less likely than Caucasians to report a hallucinogen use disorder. For past-year diagnoses, 24.1 percent of the sample met criteria for at least one SUD. The ethnic distribution for past-year use was similar to lifetime use. Within most of the public sectors examined, non-Caucasian youths were less likely than Caucasian youths to have a SUD but were more likely to be found in more restrictive service settings, including juvenile justice, where they may be less likely to receive appropriate services for alcohol, drug, and mental health problems. Although efforts are being made to address this type of disparity, more effective assessment and interventions should be incorporated across service sectors for multi-problem youths with SUD's. A discussion of study limitations suggests that these results may not be generalized to other regions of the country. 5 tables, 2 figures, and 52 references