BACKGROUND: Globally, shifts in United States foreign-aid policy have put HIV funding under duress. South Africa is unique in the region because its HIV program is largely domestically funded, although donor funds and partnerships support key components. We estimated the potential epidemiological impact of ceasing provision of services currently supported by PEPFAR and the costs and cost-effectiveness to the South African government (SAG) of potentially taking over these services. METHODS: We used a costed version of Thembisa, a South African HIV transmission model, to simulate four scenarios: a minimum scenario assuming intervention coverage reducing proportional to PEPFAR's funding share of specific activities in 2023; a maximum scenario assuming additional health system impacts; and sub-scenarios either with 3-year recovery (2029-2031) or no recovery to previous coverage. HIV program costs were estimated from the provider perspective (SAG) in 2024/25 US dollars. RESULTS: Over 2025-2028, discontinuing activities currently funded by PEPFAR in South Africa without replacement by SAG would result in 150 000-296 000 additional new HIV infections (29-56% increase) and 56 000-65 000 additional AIDS-related deaths (33-38%). Permanent discontinuation of currently PEPFAR-supported services over the next 20 years increases this to 1.1-2.1 million additional new HIV infections and 519 000-712 000 additional AIDS-related deaths. Sustaining these services would cost an additional $620 million to $1.4 billion between 2025 and 2028. Under a reduced budget, the most cost-effective interventions to preserve are ART and PrEP for key populations. CONCLUSION: Unmanaged PEPFAR exit from South Africa threatens to undo a decade of progress unless services are taken over by other funders, including SAG.