Drawing upon a nationally representative survey sample of African American (AA) drug injectors and non-injectors, this study tests for a suspected causal association between dropping out of school and the occurrence of injecting drug use (IDU), which remains a major cause of human immunodeficiency virus (HIV) transmission in this population. The data are from public use files of the National Household Surveys on Drug Abuse (NHSDA) conducted between 1991 and 1995. From within the NHSDA's nationally representative sample of adult household residents, a total of 389 AA adults with a history of IDU were matched on neighborhood of residence with 2253 AA adults with no history of IDU. The conditional form of multiple logistic regression was used to estimate the relative risk of having injected a drug for school dropouts relative to a reference category of AA who received the high school diploma but did not go to college. AAs who dropped out of high school were an estimated two times more likely to have injected drugs. With statistical adjustment for age, sex, and Hispanic background, the estimated association was 1.9 (95% confidence interval (C.I.) = 1.3-2.6, P<0.001). Contrary to our advance hypothesis, earning the graduate equivalency certificate (GED) did not seem to affect the magnitude of excess risk for having started IDU (adjusted odds ratio (aOR) = 2.3, 95% C.I. = 1.4-3.8, P<0.001). Hence, school dropout prevention might reduce the risk of IDU per se, in addition to the many other general benefits of educational attainment. The issue of GED-associated reduced risk of IDU remains open for future study.