The population-level prevention benefits of HIV PrEP will strongly depend on how the medication is used by those at substantial risk of infection. Questions remain about the potential of PrEP to achieve long-term reductions in HIV incidence among MSM given uncertainty about patterns of uptake, adherence, and discontinuation of the medication; these patterns will likely differ from the CDC’s clinical practice guidelines on PrEP indications and utilization. Deviations include: 1) the use of PrEP by MSM who self-select in using the medication but have no (or minimal) risk indications; 2) the non-use of PrEP by MSM with strong risk indications, based on either lack of awareness of PrEP or limited access to PrEP providers; 3) time-varying patterns of adherence (or non-adherence) to PrEP relative to time-varying patterns of high-risk sexual activity, compared to daily use recommended by CDC following FDA approval guidelines; and 4) heterogeneity in how MSM fully discontinue PrEP use unrelated to their ongoing risk indications, based on clinical, biological, or psychological reasons in stopping medication. In this study, we propose to build on the robust HIV PrEP implementation in our prior network-based simulation models to address questions of how longitudinal variation in these various factors could impact both the epidemiological impact and cost-effectiveness of HIV PrEP as a prevention intervention for MSM in the United States.