Research has demonstrated important linkages between crack smoking and HIV/AIDS, especially through increased risk of sexual transmission. Among HIV prevention interventionists and researchers there has been considerable speculation regarding the potential for HIV transmission through the sharing of crack pipes or stems contaminated with blood from smokers' cut and burned lips, or from oral sex performed by crack users with injured lips.' We previously found that some crack users also believe that the way crack is smoked may increase the probability of lip injuries and HIV transmission through exposure to blood on shared crack-smoking equipment (Porter and Bonilla, 1993). Although epidemiological investigation has not yet documented the relationship between injuries to the mouth caused by smoking crack and the transmission of HIV infection, there is a theoretical possibility for transmission of HIV through exposure to blood and semen on injured lips, though it is not clear how great this risk may be. There is also the possibility that lip injuries and sharing of crack stems might transmit other bacterial and viral infections. In this exploratory study we systematically address crack smokers' evaluation of their risk of acquiring HIV from injured lips. We also investigate the prevalence of these behaviors perceived as possible risk factors for HIV infection-specifically, the relationship between the way crack is smoked and the extent of lip injuries; the relationship of lip injuries to the sharing of crack pipes and unprotected oral sex; and the demographic variables associated with crack-smoking practices that might potentially place the user at risk for the transmission of HIV or other diseases. The use of crack cocaine has increased dramatically since 1986 (Abramowitz et al., 1990; Mieczkowski, 1990). In Philadelphia, Pa., the city we utilized for our research, there has been an epidemic of crack cocaine use; cocaine is the primary drug of abuse by individuals who enter drug treatment, accounting for 61% of treatment admissions, and crack-smoking among clients of city treatment programs rose dramatically from 1987 to the 1990s (City of Philadelphia, 1992). Crack cocaine is a powerful and rapidly acting drug with a fleeting rush, typically 15 minutes or less. Unlike heroin, crack use does not have any inherent physical limitation on the extent of daily dosage. The short duration of the high results in repeated consumption, sometimes 30-40 doses or more in a 24-hour period. Repeated consumption is also precipitated by the need to escape the downward crash after the initial rush. The stimulant quality of cocaine results in extensive periods of sleep deprivation, making crack users prone to binge for several days at a time (Mieczkowski, 1990; Waldorf et al., 1991). Crack can be smoked in either a crack pipe or a (pipette or tube); a crack-smoking device may also be referred to as a stem. The of crack is placed on a screen or filter and vaporized with heat, and the vapors are drawn through the filter. When a straight shooter is utilized, the pipette is held vertically to balance the rock on the filter, which is often brass wool wadded into a ball and inserted in the end of the pipette. Although glass is the most common material for pipes or stems, there is increasing use of metal pipettes or stems, including car antennas, as devices with which to smoke (Waldorf et al., 1991). Crack is sold in vials or caps in Philadelphia, generally costing $5.00 for two caps. After running out of money for crack, smokers can continue getting high by scraping resin off the side of the stem, where it accumulates during the process of smoking crack, and reheating it on the screen (Ratner, 1993). Typically this is accomplished by using an umbrella strut as a pipe cleaner, often chipping the glass ends and leaving a sharp surface (Curtis, 1994). Several studies indicate that there is an epidemiological connection between crack-cocaine addiction and HIV infection (Inciardi et al. …