PIP: AIDS cases attributable to using contaminated needles in nonmedicinal drug injection have increased in Argentina from 11.3% in 1987 to 39% in 1991. A similar increase (from 1.8% in 1985 to over 30% in 1991) occurred in Brazil. To complement existing information, data searches were conducted and personal communications from current researchers were collected for a total of 24 documents from Argentina and 18 from Brazil. The median sample size was 68 in Brazil and 188 in Argentina; most studies were from outpatient facilities, males constituted more than two-thirds of the sample in half of the studies, and the median age (when reported) was between 16 and 29 years old. Analysis of data from selected studies showed that HIV seroprevalence among drug injectors in both countries has increased rapidly, with the greatest increase occurring in Brazil. This rapid increase may also be influenced by the fact that cocaine, rather than heroin, is the drug of choice. Cocaine injection involves drawing blood into the syringe before injection and also more frequent injections. The level of seroprevalence among drug injectors varies among different subpopulations. Sexually transmitted disease clinic attenders who are IV drug users show rates of 6.58% and 51.9% seropositivity. Incarcerated persons who are drug injectors were associated with seropositivity rates of 35% (adults), 60.9% (adolescents), 53.5% (adolescents in security institutes), and 18% (female inmates who volunteered for testing), with drug injection the most important risk factor in 90.4%. Among prostitute injectors, rates were reported of 20% in 1988 and 50% in 1989-90 in the same population. Street children in Rio de Janeiro who use IV drugs (68 of 3389 surveyed) had a rate of 13.2% in 1987-88. The risk factors associated with HIV infection among drug injectors are socioeconomic status and injecting and sex practices. Co-infection patterns among drug injectors have also been found, with concomitant HIV-1 and HTLV-1 reported in 20% of 85 HIV positive drug injectors. Overall, these data suggest that there is an absolute increase in the number of AIDs cases among drug injectors rather than a decline in the proportion of cases reported in other transmission categories. Drug injector transmission poses a threat to the sex partners and offspring of injectors and provides a bridge to the heterosexual infection of women. While there is no obvious quick solution to the problem of IV drug use, there are programs which can slow the spread of HIV among injectors. It is urgent to control sexual transmission and drug injection transmission of HIV in these countries.