In this study, a microsimulation model is used to assess the potential impact of condom use on women's lifetime risk of acquiring HIV in rural southern Malawi. The model draws on survey data for information on sexual activity, marriage and divorce, and on the biomedical literature for input parameters governing the transmission and spread of HIV and other sexually transmitted diseases (STDs). We show that lifetime risk could be as high as 42 percent with no condom use and as low as 8 percent if everyone consistently uses condoms with nonmarital partners. Next, we examine the impact of more realistic, intermediate strategies of condom use, varying men's propensity to use a condom with nonmarital partners, varying the per-coitus probability of condom use, varying probabilities of slippage or breakage, and finally, examining the effect of condom use in the presence of STD symptoms. We demonstrate profound effects of consistent condom use and of condom use prompted by symptomatic STDs.