We developed a mathematical model using a multiple source of infection framework to assess the potential effect of the expansion of highly active antiretroviral therapy (HAART) coverage among those in medical need on the number of individuals testing newly positive for human immunodeficiency virus (HIV) and on related costs in British Columbia, Canada, over the next 25 years. The model was calibrated using retrospective data describing antiretroviral therapy utilization and individuals testing newly positive for HIV in the province. Different scenarios were investigated on the basis of varying assumptions regarding drug resistance, adherence to HAART, therapeutic guidelines, degree of HAART coverage, and the timing of HAART uptake. Expansion of HAART lead to substantial reductions in the growth of the HIV epidemic and related costs. These results provide powerful additional motivation to accelerate the roll out of HAART programs aggressively targeting those in medical need, both for their own benefit and as a means of decreasing new HIV infections.