OBJECTIVES: Many therapies have immediate costs but delayed benefits. Recent and anticipated transformative therapies may exacerbate these challenges. This study explored whether disconnects between short-term budget impacts and long-term costs and benefits, and among impacts on initial payers, downstream payers, and society, are expected for a range of such therapies and whether they are likely consistent or variable, with implications for potential policy responses. STUDY DESIGN: Modeling. METHODS: We modeled the impacts of 5 hypothetical therapies affecting different patient types: curative gene therapy for a childhood disorder, highly effective hepatitis C virus therapy, disease-modifying Alzheimer disease therapy, and cardiovascular disease therapy for both rare genetic and higher-risk prior cardiovascular event populations. We constructed disease-specific models, modifying best-available Markov analysis estimates for standard-of-care state transition rates, utilities, and costs. We disaggregated total healthcare impacts into impacts on initial versus downstream payers, dividing payers into 3 types: commercial insurers, Medicaid, and Medicare. RESULTS: Although we found gaps between the impacts on initial and downstream payers in all examples, some substantial, the magnitude and reasons vary. CONCLUSIONS: As scientific advances generate transformative therapies with substantial structural disconnects between "who pays" and "who benefits,"creative approaches may be needed by manufacturers, payers, and others to ensure appropriate access to cost-effective therapies, adequate economic incentives for future development, and sustainable payer economics. Mechanisms may amortize high up-front costs over time, provide for transfers among payers, or a combination. Our research suggests that approaches should be tailored to specific disease and therapy characteristics to be effective.