In order to predict the level of HIV risk for the inmates, researchers created 12 explanatory variables, including gender, race, education, age, drug use, drug sales, knowing persons with HIV or AIDS, self-efficacy, perceived HIV risk, HIV knowledge, and vulnerability to HIV. The study measured behaviors known to increase risk for HIV infection such as injection drug use and unprotected sex with multiple partners; knowledge of modes of HIV transmission, risk factors, and ways to reduce risk; attitudes toward reducing specific HIV-related risk behaviors; and self-assessments of HIV risk. Chi-square tests were used to examine the relationships among pairs of the explanatory variables and to explore the bivariate relationships among each of the explanatory variables and the risk groups. Participants were classified into risk groups based on their sexual and drug-use behaviors prior to incarceration. Compared to those in the low-risk group, those in the high-risk group were more likely to have used or sold drugs and to have lower self-efficacy and perceived-risk scores. Respondents in the moderate-risk group were more likely than those in the low-risk group to be young, to have sold drugs, and to have lower self-efficacy scores. This paper discusses the implications of these differences among risk groups for HIV-prevention programs that are tailored to risk profiles. 4 tables and 42 references