The study attempts to explain the differences in terms of income, ease of acquisition, and availability of substances from friends. Data were obtained three times during the 5-year period from 847 students in 7th, 8th, and 9th grades at 11 schools in Los Angeles County, Calif. Study subjects were selected from the original sample of 1,634 which included 13 percent Hispanic, 15.3 percent black, 7.7 percent Asian, and 64 percent white. A questionnaire was administered to the 277 male and 570 female subjects the 1st, 4th, and 5th years of the study. It was hypothesized that ethnic groups have differential access to substances and economic resources to purchase various drugs. To these hypotheses, availability from friends, perceived ease of acquisition, income from earnings and gifts/allowances, and initial substance use were examined across ethnic groups, and then used as covariates of the substance differences. A split-plot repeated measures design with covariates was used to compare changes in substance use across time and between ethnic groups. Results indicate consistent and significant differences among ethnic groups substance use at all three points in time. Earned income made a significant impact on explaining the ethnic differences for cigarette, alcohol, and given income on cannabis consumption. Adding community variables such as availability from friends, ease of acquisition, and initial drug use not only eliminates the effects of income variables on drug use, but in most cases, the ethnic differences among adolescents as well. Tabular data and 33 references. (Author abstract modified).