We will show how we made 10 year predictions of the potential effect of increasing the usage of antiretroviral therapy (ART) on preventing human immunodeficiency virus (HIV) infections and on averting Acquired Immunodeficiency Syndrome (AIDS) deaths in the San Francisco gay community. We developed a new mathematical model that reflected the transmission dynamics of both drug-sensitive and drug-resistant strains of HIV. We coupled this new model with a statistical approach that enabled us to include a high degree of uncertainty in estimating the potential: (i)treatment effects of ART (in terms of infectivity and survival), (ii) increase in levels of risky behavior, and (iii) rate of emergence of drug resistant strains. We then analyzed our model by using time-dependent uncertainty analyses (based upon Latin Hypercube Sampling)in order to predict two different futures for the HIV epidemic in San Francisco. We also performed two time-dependent sensitivity analyses. Our results revealed that substantially increasing the usage of ART in San Francisco would decrease the death rate (for either predicted future) and could substantially reduce the incidence rate (for only one of the predicted futures).